Guide to Research in Music Education

A Guide to Research in Music Education Second Edition

ROGER P. PHELPS

The Scarecrow Press, Inc. Metuchen, N.J., & London 1980

.

),

*

-,-.-,-.,...

j .

*L < ., ! .

...,

?-,:

.

-

$

5

.

i;; i-...---..--.. ; , :j

.:$

--..

.-.

r:

E

,

:';.

j .' . ' i

'-. :

;

. .U..

;

I

5

;

---..-.-.--..--.

The first edition of A Guide to Research in Music Education was published by Wm. C. Brown Cornpany in 1969.

Library of Congress Cataloging in Publication Data Phelps, Roger P A guide to research in music education. Includes index. 1. Music-Instmction and study. I. Title. MTl.P.5 1980 780' .7 80-12107 ISBN 0-8108-1303-3

Copyright @ 1980 by Roger P. Phelps Manufactured in the United States of America

This book is lovingly dedicated to my wife, Mildred, and to our sons, Roger and Homer

This textbook is a practical guide to the basic concepts and techniques of research methodology as they apply specifically to music education. The increased demand for postbaccalaureate and advanced graduate instruction has led many graduate schools to add courses in research methodology in music education. Yet because there is no general agreement on the meaning of "music-education research, " these courses have varied in content both qualitatively and quantitatively. The object of this book is to present a concise, practical, and logical approach to the fundamental principles and methods of research that music educators can readily understand and apply. Because of their usual preoccupation with the development and maintenance of performing skills, these educators are often unfarniliar with the philosophical and pragmatic connotations of research. The concepts and techV

niques described in this treatise are those that researchers in vanous disciplines have used successfully. Largely an outgrowth of this writer's experience in teaching graduate students in music-education and general-education research methodology classes, this book is designed for Courses in the fundamentals of music-education research, for thesis seminars, and for independent researchers in music education or the arts and humanities. It undoubtedly will be most useful to students in either a master's- or a doctoral-degree program, but the postdoctoral researcher will also find it of value, and the adviser will be able to use it for reference purposes. At present there are several excellent publications on educational research, but their utility to music education has often been less than satisfactory to music-research tyros. It was at the urging of several of his students, who expressed the conviction that a book was needed to concentrate specifically on problems peculiar to music-education research, that the writer prepared the first edition of A Guide to Research in Music Education. The everchanging field of research created a demand for a revision of the book. The content and bibliography for each chapter have now been updated. Reflective thinking, originally included in the second chapter, has been moved to the first, where the scientific method is discussed. The third chapter contains hypothetical statements relating to various components of a research design, with a discussion of why some are better than others. A section on oral history has been added to the chapter on historical research. I have expanded and largely rewritten the chapter on experimental research, incorporating the material on statistics that was separate in the first edition. Discussion of ethnographic research and reCent doctoral studies make the chapter on descriptive research more current. The chapter on philosophical inquiry has been expanded and largely rewritten to include philosophical Systems and philosophical analysis. The former chapter on analytical research,

dea]ing with musical analysis, has been replaced by one on aestheti~inquiry, which includes the analytic process of aesthetic inquiry and aesthetic methods. A unique feature, found in no other textbook, is the chapter on writing the research report. Various footnotes are listed as they would appear in five of the most frequently used style manuals. The decrease in funding has nece~~itated considerable changes in the treatment of funding, arid some representative postdoctoral research is included in the final chapter. This treatise can be especially useful when the pnnciples and methods it discusses are translated into the formulation of an acceptable research design, with its adequate implementation, and subsequent dissemination of the research findings. This is the central theme of the book and its principal reason for being. The writer is especially indebted to, and grateful for, the constructive criticism provided by his wife, Mildred Wade Phelps, Professor of Music at The King's College, Briarcliff Manor, New York who rvad the entire manuscript. Gratitude for contributions of the following are also noted with appreciation: Mrs. Ruth Hilton, Music Librarian, New York University, for her criticism and suggestions to strengthen Chapter 4; Dean Arnold Spinner, Director of Educational Research and Field Studies, New York University, for his comments and assistance with Chapter 11; and music researchers who graciously provided information for Chapter 12. The wnter also is grateful to the many publishers, noted throughout the book who granted permission to quote from their ~ublications.Finally, a debt of gratitude is due students in my research classes at New York University and Duquesne University for their suggestions to strengthen Chapter 3.

Roger P. Phelps Baldwin, New York August 1, 1979 vii

Steps in the Formulation of a Research Problem

3,

4.

~rganizingthe Research Proposal

61

The Research Design Format for Nontraditional Doctorates Deficiencies in Research Designs

65 81 82

Obtaining Resource Materials Efficient Library Use Library Card Catalog Suggestions for Note Taking Obtaining Items from Other Libraries Photocopying Materials Reference Materials for Music

5.

Contents 1.

Research Selection of a Topic Categories of Music-Education Research The Research Climate Research Concepts The Scientific Method Reflective Thinking

2.

Formulating the Problem Sources of Information Vulnerability of Music-Education Research ~ ~ i t iConsiderations al in the Selection of a Research Problem

...

Vlll

Historical Research: A Chronicle of the Past Purposes of Historical Research Historical Research Procedures Source Materials Selection of Topic Oral History Historical Research Data Gathering Extemal Criticism or Authenticity Intemal Criticism or Credibility

The Setting for Research in Music Education

6.

48

Experimental Research: Purview of the Future hrpose of Experimental Research Kinds of Experimental Research Concepts of Experimental Research Testing Hypotheses Selection of Subjects The Pilot Study

90 91 92 101 104 105 107

121 122 124 125 127 128 130 138 143

150 156 158 161 163 164 166 170

Other Sources of Subsidization for Research in Music Preparation of the Research Proposal for Funding

Designs for Conducting Experimental Research Factorial Designs Ex Post Facro Research Basic Statistical Concepts

7.

Descriptive Research: An Account of the Present

8.

9.

Notes Index

Philosophical Inquiry: Rational Quest for Truth Metaphysics Science Versus Philosophy Purposes of Philosophical Research Philosophical Systems Philosophical Analysis Techniques of Philosophical Inquiry

.

Aesthetic Inquiry: A Search for Beauty Reflective Thinking in Aesthetics The Analytic Process of Aesthetic Inquiry Aesthetic Methods Approaches to Aesthetic Inquiry

10. Writing the Research Report Format Mechanics Style

11. Funding for Research in Music and Music Education Federally Supported Research Foundation Support for Research

323

12. Music-Education Research: Today and Tomorrow 330 Postdoctoral Research Signs on the Honzon for Music-Education Research

Purposes of Descriptive Research The Disparity of Descriptive Research Kinds of Descriptive Research Ethnographic Research Tools of Descriptive Research

322

Figures Delimitation of a Research Topic Format for a Research Design Deficiencies Reported in Research Designs Library Catalog Card Random-Numbers Table 2 X 2 Factorial Design Normal Probability Curve Positively Skewed Curve Negatively Skewed Curve Rank-Order Correlation of Scores Obtained on Music Tests Omnibus Request for Information Likert-Type Scale Plato's 'Divided Line " Process Ecker and Kaelin Levels of Discourse Dissertation Title Page L

33 1 337

1. The Setting for Research in Music Education The nse and fall of civilization is a chronicle of humanity's attempt to find solutions to problems. Human beings have not always been concerned with a formalized plan to seek their way out of a dilemma, as is frequently the case today; rather, they have often relied on the process of trial and error. Some of the greatest discovenes of the world have been the result of serendipity. The word "serendipity" is derived from the account of the Three Princes of Serendip, as told by the eighteenth-century British novelist Horace Walpole. As they sailed the Indian Ocean, these princes of Serendip (now known as Sri Lanka) were continually finding thngs they did not expect. Examples of discovery by serendipity include that marvelous antibiotic penicillin, discovered by Alexander Fleming quite by accident. X-rays, which are

Research in Music Education

The Setting for Research

2

so widely used today, were revealed in much the Same way to Wilhelm Roentgen. The telephone is generally accepted as having been discovered by chance when Alexander Graham Bell was experimenting with a device to enable his deaf wife to hear better. To approach serendipity from a hypothetical viewpoint, Sir Arthur Sullivan's fanciful "discovery" of The Lost Chord would have been the result of serendipity had the traditional bases of music theory not been so well established already that a "new" chord was an impossibility. Any significant research findings, except those that result from serendipity, are usually based on careful and deliberate planning, followed by precise execution of the plan, and then unbiased critique of the results. Frequently this cycle is repeated several times before significant results are achieved. Problems have always confronted humanity and will continue to challenge its ingenuity in the future. The continuum is endless, because as answers to questions emerge through experimen'tai and philosophical procedures, new and additional challenges arise as we move up the ladder of progress. Only when our hunger to improve our lot in life is satisfied will our quest for knowledge be fulfilled. The human race by instinct is committed to change, since without it survival is not possible for long. The Same is true in the animal realm. Witness the demise of the large prehistoric brontosaurus and the renowned dodo. Their disappearance from the face of the earth has been attributed to an inability to adjust to environmental changes. As a possible explanation of this evolutionary process, Charles Darwin engendered the principle of "survival of the fittest" in his Origin of Species, a publication that has continued to arouse controversy since it first appeared in 1859. Since Homo sapiens is regarded as the most intelligent of all creatures, it is not u~easonabieto expect that it will continue seeking to perpetuate itself by finding answers to its perplexities in an intelligent and rational way, something that many fauna apparently have been unable to do.

'

I

1

3

The process of metamorphosis has extended to the tools and devices we use not only for our survival but also for our enjoyrnent. For instance, even a cursory glance at a music-history textbook will reveal names of many instniments that are no langer in use: ancient Greek cithara, the medieval shawm, and the ophicleide of the last century, to cite a few. Reasons for their obsolescence are beyond the scope of this book, but in common with the brontosaurus and the dodo they evidently outlived their u s e f u l n e ~ ~It. might be added, parenthetically, that the current disuse of these instruments, however, is no reason for failure to ]eam about them. Any Person well read in music, as in any discipline, is expected to know about significant developments and events pertinent to the field. A knowledge of the past is essential if one is fully to understand the present. A music educator may acquire this information through research into many sources. The word "research" is used more loosely in this instance than it will be throughout the remainder of this book. Even in this illustration, however, it does imply a "seeking of information," although not necessarily in depth. The quest for solutions to problems, as already suggested, may be called research in the proper sense of the word. The sequence seemingly is endless, because change brings with it new riddles that often defy immediate resolution. Modification, then, is frequently attended by some organized type of research to cope with the problem that has arisen. Some solutions to problems obviously result from procedures that are less formally organized, because, as Kelley avers, research actually is a process of evolutim.' In the realm of music, Johann C. Denner's clumsy ejghteenth-century clarinet of five keys would be repugnant to the contemporary clarinetist, who performs on an instrument containing up to twenty keys and seven rings. One has to wonder ~ ~ e t h the e r cithara, shawm, or ophicleide likewise might have been modified and improved to keep Pace with changing aesthetic arid musical concepts. But this conjecture hardly seems appro-

Research in Music Education 4 priate: the record is clear. Obsolescence or usefulness are both the result of change, but for different reasons, many of which will never be known. It should be noted, however, that research may hasten either one or the other. From a practical standpoint, research is concemed with utility, although from a historical perspective, obsolescence would provide appropriate topics for study .

Research The term "research" has many meanings, some rather general and others more specific. Hopkins states that research is "stnictured inquiry that: 1. Utilizes acceptable scientific methodology to solve problems; 2. creates new generally applicable knowledge. "* It is this writer's contention that resgarch is a carefully organized procedure that can result in the discovery of new knowledge, the substantiation of previously held concepts, or the rejection of tenets that have been widely acclaimed. Drew reports that research "is a systematic way of asking questions, a systematic method of enquiry. "3 Engelhart also presents a definition of educational research that is applicable to music education; namely, the identification of generalizations that enable educators to determine what is necessary to do for the education of children and adult^.^ Madsen and Madsen point out that research is not only a "way of thinking" and a "state of mind," but also a way to "think rationally and objectively concerning the aspects of life that can be studied scientifically. Sound research consists of organization, execution, and dissemination, in logical sequence. Research obviously cannot be conducted properly without a good plan of operation. On the other hand, even a well-organized prospectus may not be implemented effectively, resulting in insufficient data. Both organization and execution must be of the highest caliber. In scientific

The Setting for Research

5

research dissemination is almost universally a concomitant aspect of a project. Sometimes this is not the case, unfortunately, in

research in education and in the humanities. If research in music education is to be worthy of the name, the results should be solid enough to be shared with members of the profession. It hardly seems necessary to mention that the investigator is cmcial to the success of all research. The investigator must be both critical and inquisitive, as well as be able to determine what type of research is most feasible for him or her to implement, as Barzun and Graff so clearly point out.6 To the musician, largely trained in nonverbal skills, the idea of undertaking research may seem foreign. Music teachers often fail to follow the logic of a researcher and thus neglect to take advantage of research that can provide practical answers to problems. Yet the increasing demand on the part of state departments of education for a minimum of a master's degree as the requisite for permanent certification has resulted in a phenomenal growth in graduate programs in music education all over the nation. While some master's-degree programs are organized to allow the student automatically to cany extra course work in lieu of a written project, there are others that mandate some type of creative endeavor (composition, arrangement, thesis, recital, or field project) as partial fulfillment of degree requirements. The option of either extra course work or a project exists in some master's-degree curricula, a discretionary arrangement that has some advantages. n e experience of this writer in advising students at both the master's and doctoral levels, however, suggests that graduate schools should consider the feasibility of making some type of culminating written requirement mandatory for all master's matriculants. Not only can the students expand their knowledge and receive intellectual stimulation, but they will have had some practical research experience should they eventually pursue their doctorates. The demand for academic standing above the baccalaureate

Research in Music Education

6

degree began in the 1920s. Until the advent of World War I1 the master's degree generally was considered to be adequate for music teachers, even on the college or university level. In the past thirty years, however, the pressure on college and university music teachers by administrators has caused a rethinking of graduate music education, with the result that many beginning postbaccalaureate students now anticipate that a master's degree may only be a transitional step to a terminal degree. An aspirant for a teaching position at the college or university level now can hardly hope to rise above the rank of assistant professor without an earned doctorate except in a few isolated situations. A byproduct of this coercion has been the demand for doctorates with emphases different from those the traditional Ph.D. and Ed.D. have presented. Several colleges and universities in the United States now offer the D.M.A. (Doctor of Musical Arts) and the Mus.D. (Doctor of Music). A recent development is the D.A. (Doctor of Arts), a more flexible, interdisciplinary, and individually oriented program of studies that enables a student to relate music to other arts more effectively than can either the traditional Ph.D. or Ed.D. The end product of the D.A. is a research paper that is usually more product- or process-oriented than other doctorates. For example, one of this writer's D.A. candidates has been examining the metallurgical properties of mouthpieces and lead pipes of some brass instruments, using sophisticated electron-scanning microscopes in an engineering laboratory. There has been considerable criticism, much of it justified, of the so-called "nontraditional doctorate, " which is interpreted to be other than the Ph.D. or Ed.D. Although the D.A. is a new degree, it should not be placed in the Same class as the nontraditional doctorate. The D.A. programs with which this writer is familiar, unlike some of the nontraditional doctorates, screen candidates carefully, have certain residency requirements, and maintain high academic standards.' Since all earned doctorates at present are based upon the acceptance of some type of "creative" project, students in music edu-

The Setting for Research

7

cation as well as other disciplines find themselves faced with the reality of engaging in research. Many times selection of an acceptable topic becomes an unnecessarily long and tedious process, a process that might be shortened considerably if the student were to use reflective thinking to help locate and develop a meaningful and interesting topic. The necessity for researchers to do their own thinking is emphasized by Koefod, who also observes that an excellent research report is the hallmark of this kind of investigator.'

Selection of a Topic Fortunately, many students give considerable thought to potential projects prior to enrollment in a graduate program. Others depend upon expediency, mandate, suggestions from an adviser, or some other extrinsic factor in choosing a topic. Such ambivalence can hardly result in anything more than inferior productivity. Researchers should choose a topic for investigation in which they have an intense interest, one to which they may lay claim as "their own. " It is inconceivable that anyone would become involved in a research topic in which he or she is neither totally engrossed nor in general agreement philosophically. School administrators recognize that personnel perform most efficiently in the areas where they are most competent and interested. In the interest of educational efficiency, good administrative practice dictates that such predilections be honored wherever possible through appropriate assignments. Why should involvement in research operate under a different procedure? As if in reply, Kelley notes that the conduct of worthwhile research is not easy, and he also deplores those who dismiss its importance too l i g h t l ~The . ~ selection of an appropriate topic is very signifiCant for the researcher, and it will be treated more comprehenin the next two chapters. While most graduate students pursue research in music educa-

Research in Music Education

8

tion for pragmatic reasons, a gratifying trend is evident in the ever-increasing number of postdoctoral research titles. Govemment, philanthropic agencies, or foundations have supported many of these projects. Unfortunately, much of the money allocated by these agencies in the past has dried up, especially for the arts. These projects were subsidized, moreover, only after a very careful review of the research procedures proposed. The experience received in developing and pursuing a graduate project ought to make it easier to prepare an acceptable proposal for funding from agencies and organizations that still support research in the arts. On the other hand, it would not be accurate to infer that all research plans that are rejected for funding are deficient in organization. Many excellent proposals have not been subsidized and will not be unless more monies become available in the future. The quality of research in music education, like that of education in general, has been Open to question in many quarters. Some of this criticism is justified. Sometimes a research study consists of a superficial treatment of some insipid topic that the busy music educator did not really find pertinent. Actually, there are innumerable significant problems in music education that need to be solved. Might it not be more useful to concentrate first on those that are of immediate concern and practical value to the profession? When realistic solutions have been found to these, then some attention may be given to those areas that are of less import to music educators. Realistically, it must be admitted that some of the most crucial areas do not lend themselves to easy and quick solutions. The time schedule of a researcher and financial subsidization are frequently both limiting factors. Since times change, as indicated earlier, some solutions may neither be found nor needed for certain perplexities. One might ask, why engage in research if it will not be beneficial or if the results will be inconclusive? Significant research rarely results when answers to problems are obtained in haste. In addition, such findings may even be erroneous or impractical note Borg and Gall.lo

The Setting for Research

9

A factor that has affected both the quality and kind of research conducted in music education is the dichotomy to which this writer called attention in 196011:on the one hand, there are those educators who feel that music is a practice or skill and as such does not lend itself readily to research techniques. This group is opposed by those who recognize that music can and does properly adapt itself to the recognized criteria for research but who feel that it has not been conducted in as much depth as desirable. The recent encouraging increase in research projects undertaken by musicians and music educators suggests that at last the climate for research activity has changed. Even a casual glance at titles of some of the research completed in music will substantiate this optimistic observation. A recent listing of research sponsored by the U.S. Department of Health, Education, and Welfare reveals that investigators in music have taken many directions.12 As music researchers have become better trained there has been a tendency to overemphasize behavioral approaches to research. Many answers can best be obtained with experimental procedures; however, some questions can better be answered by using other methodologies, as this writer points out in a recent article.13 The careful structuring of research projects in music education, regardless of methodology used, should be accompanied by a subsequent meticulous implementation of the research plan to bring about valid and objective results that, when disseminated, will be meaningful and practical to members of the Profession.

Categories of Music-Education Research Most research studies in music education fall under one of the following categories: historical, experimental, descriptive, ~ h i l o s o ~ h i c aor l , aesthetic. A purview of any Standard textbook ln educational research will likely include a discussion of the first t h e e types. Philosophical inquiry (or research) will be contained

Research in Music Education

10

in a few, but many educational researchers either deny its existence as a separate kind of research or suggest that its use is so extremely limited as to be hardly worth serious consideration. Yet in music education philosophical inquiry can be a very fruitful area for investigation. The last type, aesthetic inquiry (or research), is a designation that normally relates to discovery of the beautiful in the arts. In music, however, it rnay relate to learning about a composer or the composer's music, or even to any other area of investigation that focuses on music. When concerned with the study and analysis of certain compositions of a Composer, including the latter's role in the mainstream of music history, it rnay be regarded quite properly as quasi-musicological in scope. Many possibilities exist in the aesthetic type of research for the music educator who has a good theoretical and musicological background. Since the techniques, procedures, and characteristics of each of these kinds of research as they relate to music education will be presented in detail in subsequent chapters, they will just be mentioned here. Any one study, of Course, rnay incorporate some aspects of the other research types, but its emphasis will be on one of the five areas just listed. The organization of a study largely detennines the format under which the research will be conducted. A historical study will emphasize the design for a historical study, a descriptive study the aspects of that kind of project, and so on. The labels associated with a given research study should not become a fetish, however, because the essential factor is the information included in the study. Categorization, on the other hand, is helpful in determining procedures that will be or have been followed in a research study. For example, a study rnay be aesthetic-descriptive, aesthetic-historical, or any other combination. In the former, a music or other aesthetic topic will be approached using descriptive techniques, and in the latter, historical procedures . Just as most research studies are not constituted exclusively in

The Setting for Research

11

one area at the expense of the others, so, by analogy, no person is completely an introvert or an extrovert. Personality traits labe1 him or her as being inclined in one direction or the other. To cite another example, a musical tone that an observer identifies as A may have a frequency of 220 hertz (Hz.). When the sound is analyzed by an oscilloscope it rnay be found also to possess minor amounts of energy present for E (660 Hz.) or C# (1,100 Hz.), yet the listener perceives only the note A because it is the dominant sound. Likewise, an experimental study could include both historical and philosophical concepts, yet be regarded as basically an experimental study. Regardless of the kind of organization, the research must meet certain rigid criteria if it is to be sound. Many of these are totally dependent on the attitude of the researcher, not on the organization, because even the best-organized research plan is virtually meaningless unless it is implemented carefully by an investigator who uses good research techniques.

The Research Climate If music-education research is to gain and maintain the respect of the academic community, the desire for an advanced degree or promotion to a higher professional rank does not provide the right kind of motivation. As noted earlier, the investigator is usually the key to the ultimate success or failure of a research project. 1s it not logical, then, to expect that researchers be in the proper frame of mind while planning, executing, and reporting their investigations? This mental condition rnay be called the "research climate," because it implies that the researchers will involve themselves wholeheartedly in the tasks before them. Certain traits or characteristics are essential and need either to be present or to be developed if a researcher is to complete a project successfully . Ideally, one person should possess all of them, but in reality each person has his or her own limitations. While by no

Research in Music Education

12

means exhaustive, the following ten adjectives suggest the characteristics that successful investigators in music education need to possess. They should be inquisitive, perceptive, objective, discriminative , impartial, candid, diligent , persistent, creative, and erudite. The inquisitive mind is the first requisite for research. Without a passion on the part of an investigator for uncovering or learning new knowledge, research is not possible. Once the desire to find answers to problems has been affirmed, a perceptive mind is necessary to formulate the most effective plan to be used for the duration of the research. This format, obviously, should not be immutable, but subject to modification as the need for emendation arises. After the research plan has been established, investigators need to obtain information with an objective viewpoint. They must be willing to accept the results of their research even though the results may prove to be contrary to what they have believed up to this point. Researchers should be discriminative in determining what material will be useful in a written account of the results of their research. Discrimination also implies that investigators ought to be impartial and not biased in their reporting. They should avoid the tendency to weight their evidence because of some personal prejudice. A candid accounting, of Course, suggests that investigators must be forthright and frank with their presentation. This means that they accept and report accurately all their research findings. If research is tobe successful and complete, diligent attention to all details is necessary. Yet diligence alone is not enough to guarantee good research. A diligent researcher also needs to be persistent. Persistence implies that one will cany on a project despite obstacles that may either appear to impede Progress or possibly intimate that the results may be negative. Skillful researchers also should be creative. They must not only organize their project in an original way, but also execute the study and disseminate the results in a manner that is refreshing and different. Finally, researchers ought to be

The Setting for Research

13

erudite. Once the research is complete investigators should be more knowledgeable than they were before engaging in the research. In addition, they will very likely have accumulated some basic knowledge in the several disciplines that go into the making of an "educated" Person, sometimes referred to as "basic general education " or "general cultural education. " Music graduates are often handicapped in this regard, especially if they have a conservatory training, where skills are emphasized at the expense of general culture. This writer has found, as a result of advising doctoral students for several years, that this "narrow" concept often is a handicap in a doctoral program. Such students, by inclination or training, sometimes do not have the intellectual foundation to organize and execute a doctoral project. When institutions that prepare students at the undergraduate level fail to provide cumcula that train future musicians not only professionally but intellectually, they must share some of the blame when their graduates experience difficulty in completing postbaccalaureate work . Hopkins poignantly synthesizes the nature and purpose of research in these words: "A study of the evolution of music reveals that Songs, too, can be thought of as answers to questions in the minds of people. Studies of literature and music from all eras can be viewed as attempted answers to the questions that were on people's minds in each of those times. ' q 4

Research Concepts The researcher in music education can expect to encounter certain terms or concepts that relate to procedures of research. Those that refer only to specific areas or types of research will be discussed in the appropriate chapters in this book. Others, which are common to all types of research, follow. An understanding of these items is important to the logical organization of a study.

Research in Music Education

14

When matriculated, especially at the doctoral level, candidates often are required to make a distinction in their programs of studies between an emphasis that will enable them to prepare a project of pure research and one that may be called applied. The former, sometimes known as "basic" or "fundamental" research, is based on the accurate reporting of results without any attempt to incorporate practical applications of the findings in the study. Those who read the report are expected to make whatever use of the data seems appropriate to their own situation. Pure research, often concerned with the development of a theory or model based on previous theories or models, is desirable in almost all fields of endeavor, but frequently is not the type that will produce answers to the questions facing music educators in the classroom. Applied (sometimes called "action") research is conducted and reported in such a way that the investigators include practical suggestions for applying data of theirstudies to a teaching situation. Some universities permit a variant of the Ed.D., referred to as the "Alternative Ed.D.," which Stresses the development of a product, a series of position Papers, or the completion of a textbook, for example. This type of project is more flexible in format than that for the traditional Ph.D., D.M.A., Mus.D., or Ed.D., and the final document is similar to the one produced for the D.A., referred to earlier in this chapter. An example of an Alternative Ed. D. topic would be: "Three Position Papers on the Teaching of Music Methods to Elementary Education Majors in a College or University. " Sometimes the Ph.D. is erroneously referred to as a superior "research" degree because of its emphasis on the "basic" concept of research, and the Ed.D. as a somewhat inferior "teaching" degree, in which investigators are expected to indicate explicitly how the results of their research may be utilized. The elevation of one degree over the other is unfortunate. It must be admitted, however, that in some graduate programs the Ed.D. really is inferior to the Ph.D. due to differences

The Setting for Research

15

in acceptable minimum standards between them. The result is that not as much is expected from academic, musical, or research standpoints from the former as from the latter. Both types of degree programs should be predicated on sound research procedures. It is the emphasis that ought to be the essential difference between the two, not the academic or intellectual demands on the investigator. A history of eighteenth-century English secular choral music, for instance, most likely would be regarded as pure research, providing the study was concluded according to the concepts of "basic" research. On the other hand, an anthology of eighteenth-century English secular choral music could be classified as applied, because an anthology, by implication, may consist of compositions that have been collected and edited for a specific group, with rehearsal suggestions also incorporated in the study. The anthology quite properly needs to contain a history section, but this factor alone is not important enough to change the character of the study, since the basic intent in this type of research is to present practical suggestions for using the results. Of the five kinds of research enumerated in this chapter, historical, experimental, philosophical, and frequently aesthetic usually fall under the heading of pure research, whereas descriptive and some aesthetic studies might be described as applied. It should be noted, however, that any attempt to assign arbitrarily any h n d of research to one category or the other is risky at best. The data and the purpose of a research study ought to determine how the study is classified. Pigeonholing can be very hazardous, but researchers who, during the process of plaming, consider their research with either a pure or an applied emphasis in mind will find their projects much easier to organize. The terms qilalitative and quantitutive refer to procedures for looking at information. In the former, research results are largely subjective, that is, not easily translated into scores, whereas in the latter, objective data based on the scores of some type of measurement are given. Travers points out that qualitative generaliza-

Research in Music Education

16

tions eventually may lead to pronouncements that are substantiated quantitatively.16 Many research studies contain both qualitative and quantitative data. On the other hand, while it is possible for a research project to be so organized that it would contain only qualitative data, it would be virtually impossible for a study to contain quantitative but not qualitative data, because even in a study where data are the result of some type of measurement, narrative is necessary to explain the theoretical rationale for the study, and a rationale Statement is a type of information. It is not uncommon to examine research studies in music education that contain qualitative data only, with quantitative data omitted, when there is every indication that they should have been included. Such studies are ineffective and weak because the investigator did not proceed thoroughly and carefully to the next significant step of quantifying information. For example, a qualitative study might be undertaken to l e r n which school systems in a given state have instrumental music programs. These data, readily obtainable by a questionnaire or other means, could be useful to school administrators who do not have an instrumental program, enabling them to report to their Boards of Education that such a program is needed if children in the community are to receive the Same cultural advantages as Youngsters in other school systems. To music teachers, by contrast, these data would be relatively useless because they give no indication of the actual content of the instrumental programs surveyed. They merely signify the presence or absence of instrumental music in the school systems investigated. Music supervisors more likely would be interested in determining how their curricular offerings compare with those of other schools. In other words, they might be interested to l e r n that a certain school has two orchestras, two bands, or two choruses in the senior high school and thus try to convince their superiors that this Pattern should be emulated. Since music is largely concerned with skills, still unanswered is the question of performing competency of the groups under con-

The Setting for Research

17

sideration. This problem also is quantitative, and answers would be dependent on the use of a specialized type of measurement in conjunction with the questionnaire. Such an investigation would be proper under certain conditions. The investigator would need to establish criteria for comparisons in order to ascertain how well the groups performed. Researchers who organize their projects so as to obtain quantitative data are in a better position initially to produce a significant research study than are those who merely scratch the surface by seeking only qualitative information. It should be made clear, however, that both quantitative and qualitative studies may produce valid information although different techniques are used for each. It has been gratifying to note that the quality of research in music education has improved markedly in the last decade. Much of this improvement is due to greater sophistication in the field of research.

The Scientific Method Precepts of the scientific method, which is a way of thinking, were initiated and originally utilized by researchers in the natural sciences. These principles now have been applied to research in almost all disciplines. Unfortunately, music educators at times feel that the scientific method should be used only with certain types of research. This misconception evidently is due to a misunderstanding of what the scientific method connotes. Each of the five types of research enumerated earlier in this chapter should utilize the scientific method with the understanding, of Course, that some modifications will be necessary. Stated in its simplest terms, any investigation thar is logically organized, objectively implemented, und precisely interpreted meets the conditions of the scientific method. Almack succinctly defines the scientific method as "the expert pursuit of knowledge. "I6 Brennan points out that even in science, reasoning and

Research in Music Education

20

for embouchure development that result in a significant improvement in tone quality, but they still may be unable to analyze why this transformation took place. Some music educators will argue that development of practical skills should come first. Then, if there is time and inclination, philosophical and theoretical concepts may be introduced. In the natural and physical sciences, where reflective thinking and an "intellectual" approach normally are an adjunct to instruction, the scientific method is not Strange to graduate students who initiate a research project. Observing this desirable fusion of skills and theory, Good states that "science without philosophy is blind, while philosophy without science is empty . Music is a skill; consequently, the graduate student in music education who has spent countless hours in the practice room often is intellectually handicapped. Yet if research is to be successful some intellectual activity is necessary'. This intellectual process is referred to as "critical" or "reflective thinking. " In a recent article this writer stated that "critical or reflective thinking is associated with logic and the scientific method, since all these terms relate to obtaining information through metempirical rather than empirical means. "23 Metempirical is a philosophical term that refers to obtaining information solely through reasoning, whereas observation and measurement characterize the empirical method. 'lZ2

Reflective Thinking An investigator must use reflective thinking to formulate a problem into a pliable and workable format. As indicated earlier, this initial phase of research unquestionably is one of the most important, yet it frequently is treated perfunctorily. Several educational researchers, including Mason and Bramble, aver that the

The Setting for Research

21

&imate success of a research project is often related to the way the problem is f ~ r m u l a t e d . ~ ~ Concepts of reflective thinking may be traced as far back as the deductive method used by Aristotle and other early Greek thinkers. Deduction is reasoning from the general to the specific (or particul~).This type of reasoning marked one of humanity's earliest attempts to think through problems. As an example of deduction, note the choral conductor who hears some unusual sounds coming from the group and tries to determine what causes them. It turns out that the altos were singing E b , not Eh; the tenors forgot to sing F#; and the accompanist was one measure ahead of the group. Deductively, this choral conductor has observed problems and then been able to correct them by identifying specific weaknesses. In the research proposal, to be discussed in Chapter 3, the process of delineating a general problem Statement, and then forrnulating specific components, or subproblems, is analogous to deductive reasoning. Syllogism is another example of this Greek concept of deductive problem solving. In syllogism the formula of an argument consists of three propositions. The first two, known as major and minor premises, are assumed to be true, and lead to the third proposition, known as the conclusion. No attempt is made to prove or disapprove the major and minor assumptions. An example of syllogism follows: Major premise: All musicians are talented. Minor premise: Conductors are musicians. Conclusion: Conductors are talented.

In examining the logic of this syllogism one assumes that the musicians concerned are talented or they would not be able to be musicians and that the conductors could not conduct unless they had had musical training and were musicians. The conclusion

Research in Music Education

22

then is obvious. Characteristic of any syllogism is the identification of three factors, each of which is repeated twice. In the aforementioned example, "musicians, " "talented, " and "conductors" each appears two times. Of course, because of the acceptance of faulty assumptions, some syllogisms do not result in valid conclusions. Consider this example:

The Setting for Research 23

~

Major premise: A trombone is a brass instrument. Minor premise: A brass instrument has valves. Conclusion: A trombone has valves.

The major premise, of course, is true; the minor premise is only partially true. Most brass instruments contain either piston or rotary valves; however, the trombone (with the exception of the virtually obsolete valve trombone) does not. The conclusion, therefore, is not tenable. There are many classic anecdotes about those who dared to use logic different from this deductive method and who were repnmanded severely as late as the seventeenth century. For example, the announcement by Galileo that while searching the heavens with his new telescope he had discovered four moons revolving around the planet Jupiter was received skeptically by many of his Peers. One fellow professor even stated that since Aristotle had not mentioned these moons, they could not possibly exist. Others declared that since the moons were not visible to the naked eye they did not exist at all. Such was the stranglehold the Greek system of Aristotelian logic had on scholars up to the late sixteenth century, the time of Francis Bacon, who disagreed with the prevailing concept of blindly accepting deductive theories merely because they were passed on as truth by the authorities. This uncertainty about the truth of a generalization led Bacon to develop the type of thinking known as induction, or reasoning from the specific (or particular) to the general. It was assumed that reasoning based on specific items would result in more valid

1

' I I

generalizati~ns.On the basis of these specifics, generalizations could then be made about similar or related but unobserved facts or events. This type of reasoning is used occasionally by music teachers. For instance, a band director may conclude that a group will perform badly after observing several specific deficiencies before the group starts to play. First, it may be evident that the oboists do not have their reeds completely inserted in the casing, which will result in intonation that is consistently flat. Some of the French horns obviously are using the E b slide even though the music is written for F horn. The percussion section may have the wrong music in front of them. The example is absurd, of course, but it illustrates that this music teacher has observed specific deficiencies and then was inductively able to generalize that an unsatisfactory sound will result if the musicians perform under the conditions just described, even though they have not yet played one note of music. Researchers who are formulating a research design may have in mind specific components to be researched, but then they must inductively derive the general problem statement, or the gestalt. More recently research has been predicated on a combination of the two concepts; this is commonly referred to as the "deductive-inductive process," but is also known by the term '< general-to-specific-to-general." Charles Darwin is generally acclaimed to be the first to combine successfully the ideas of Aristotelian deduction with those of Baconian induction. According to Best, Darwin, in stating his theory of the ongin of species, formulated thmugh deductive-inductive procedures a process that . ~ simple ~ now serves as the basis of the scientific m e t h ~ d A illustration: after hearing a beginning string class perform badly , a music teacher analyzes the performance of the students, inductively observing deficiencies and correcting them when necessary. Finally, as a result of the inductive diagnosis of each student's performance, the teacher may deduce what will result if changes are not effected. Ary, Jacobs, and Razavieh indicate that

Research in Music Education

24

Darwin's approach, the deductive-inductive process, is in accord with the scientific method, the most appropriate procedure for .~~ of this deductive-inductive obtaining i n f o r m a t i ~ n Application process will place investigators in a much more favorable position to formulate adequately and then pursue their topics than they would without it. This concept of logic makes it relatively easy to organize research reflectively. The large number of textbooks on research methodology that use as the basis for reflective thinking the five steps originally proposed by John Dewey (1859-1952) signifies the high regard with which research leaders in virtually all areas of education generally hold this distinguished educational philosopher. Dewey's five steps, as paraphrased, are: (1) recognition of need, (2) isolation of the problem, (3) postulation of solution, (4) accumulation and codification of data, and (5) confirmation and experimental substantiation of hypothese~.~'

To these original five, other educational philosophers, such as Kelley, have added another-appraisal of the solution in light of future n e e d ~ Not . ~ ~only do these six precepts form the basis for the delineation of a problem, they also may be employed for implementation of the research. In view of current emphasis on disseminating research findings, resolution and completion of the research suggests an additional step, dissemination of the data, which might be proposed as a seventh step. Attention was called earlier in this chapter to deficiency in this regard in much musiceducation research. Many excellent reports are of no value to the profession because investigators have been either unable or unwilling to share their results. Research that is worthy of the name certainly ought to be worth disseminating. Even research of a

The Setting for Research

25

superficial nature has value for potential researchers because it may point not only to areas needing more investigation but also to mistakes and pitfalls to be avoided. These seven steps, in essence, form the basis for the forrnulation, irnplernentation, and prornulgation of a research topic. In other words, they constitute a "method" for research. The scientific method and reflective or critical thinking should not alarm a researcher in music education; rather, they need to be understood and used. Today, when so much emphasis is placed on research in music education, investigators can ill afford to proceed with a project that is not logically orgarzized, objectively irnplemented, and precisely interpreted. They may need considerable assistance and guidance to realize these objectives, especially if their undergraduate training has emphasized "skills" at the expense of the "systematic" and "intellectual" approach to learning. A college or university that does not provide intelligent and perceptive leadership to give graduate students the kind of information and stimulation they need for research is derelict in its duty to train tomorrow's leaders in music education. This does not imply, of Course, that advisers should do all the thinking and planning for their advisees, nor does it infer that any graduate students who comply with certain academic requirements will automatically receive their degrees. The determining factor should simply be whether or not an individual has been able to produce a piece of research that at least meets the minimum standards prescribed by the institution granting the degree. Students are entitled to competent advice and guidance to reach their goals. lnvolvement in research can be a most absorbing and gratifying experience, especially if an understanding and application of the concepts of reflective or critical thinking as embodied in the scientific method are uppermost in the mind of the researcher. Since the setting for research actually begins with a problem that

Research in Music Education

26

The Setting for Research

may need to be clarified or refined before it can be solved, this initial phase of research will be discussed in the next chapter.

9. What is a syllogism? Why is syllogism significant to research?

10. Name and discuss the seven steps of reflective thinking presented in this chapter.

Problems for Review und Discussion 1. What is research? Sound research involves what steps?

Supplementary Readings

2. Discuss factors important in the selection of a research topic.

3. Consider some of the problems that are of immediate concern to you, and make some type of preliminary investigation to determine whether you will be able to reach a solution to some of them.

4. Why is a proper "research climate" important for musiceducation research? What characteristics must a researcher possess in order to meet the conditions of a proper "research climate ' '?

5. What is the difference between pure and applied research? Give Synonyms for each of these terms. 6 . Differentiate between qualitative and quantitative research.

7. In succinct terms, what does the term "scientific method" imply? Why is the scientific method so important for musiceducation research? 8. Discuss reflective thinking. What differences exist between the deductive and inductive methods of reasoning? How do they differ from deductive-inductive reasoning?

27

I

ALMACK, JOHNC. Research und Thesis Writing. Boston: Houghton Mifflin, 1930, Chapters 2 and 4. ARY,DONALD, LUCYC. JACOBS, and ASGHAR RAZAVIEH. Introduction to Research in Education, 2nd ed. New York: Holt, Rinehart and Winston, 1979, Chapter 1. ASHER,J. WILLIAM. Educational Research und Evaluation Methods. Boston: Little, Brown, 1976, Chapter 1. BEST,JOHNW. Research in Education, 3rd ed. Englewood Cliffs, N.J.: Prentice-Hall, 1977, Chapter 1. BEVERIDGE, W. I. B. The Art of Scientific Investigation. New York: Vintage Books, 1950, Chapters 1, 3, 4, and 7. BORG,WALTER R., and MEREDITH D. GALL.Educational Research: An Introduction, 3rd ed. New York: Longman, 1979, Chapter 1. R., and ERNEST NAGEL.An Introduction to Logic und COHEN,MORRIS Scientific Method. New York: Harcourt Brace, 1934, Chapters 1 and 11. COOK,DAVID R ., and N. KENNETH LAFLEDR. A Guide to Educational Research, 2nd ed. Boston: Allyn and Bacon, 1975, Chapter 1. DREW,CLIFFORD J. Introduction t o Designing Research and Evaluation. St. Louis: C. V. Mosby, 1976, Chapter 1. MAX D. Methods of Educational Research. Chicago: ENGELHART, Rand McNally, 1972, Chapter 2. GAY,L. R. Educational Research: Competencies for Analysis und Application. Columbus, Ohio: Charles E. Merrill, 1976, Chapter 1. GOOD,CARTERV. Essentials of Educational Research. New York: Appleton-Century-Crofts, 1966, Chapter 1.

Research in Music Education

28

HILLWAY, TYRUS.Introduction to Research, 2nd ed. Boston: Houghton Mifflin, 1964, Chapters 1, 2, 7, and 17. HOPKINS,CHARLES D. Educational Research: A Structure for Inquiry. Colurnbus, Ohio: Charles E. Merrill, 1976, Chapters 1 and 2. KERLINGER, FREDN. Behavioral Research: A Conceptual Approach. New York: Holt, Rinehart and Winston, 1979, Chapter 1. . Foundations of Behavioral Research, 2nd ed. New York: Holt, Rinehart and Winston, 1973, Chapter 1 . MADSEN, CLIFFORD K., and CHARLES H. MADSEN,JR. Experimental Research in Music. Raleigh, N.C.: Conternporary, 1977, Chapters 1 and 2. MASON,EMANUEL J., and WILLIAMJ. BRAMBLE.Understanding und Conducting Research: Applications in Education und the Behavioral Sciences. New York: McGraw-Hill, 1978, Chapters 1 and 2. MOULY, GEORGE J. The Science of Educational Research, 2nd ed. New York: Van Nostrand Reinhold, 1970, Chapters 1, 2, and 3. PETZOLD, ROBERTG. "Directions for Research in hlusic Education, " Music Educators Journal, January 1964, 50,5:39-42. PHELPS,ROGERP. "Critical Thinking: A Prerequisite for All Sound Research, " The New York Stute School Music News, March 1978, 41,7:31-32. RUMMEL, J. FRANCIS. An Introduction to Research Procedures in Education, 2nd ed. New York: Harper and Row, 1964, Chapter 1. SAX,GILBERT.Empirical Foundations of Educational Research. Englewood Cliffs, N.J.: Prentice-Hall, 1968, Chapters 1 and 2. TRAVERS, ROBERT M. W. An Introduction to Educational Research, 4th ed. New York: Macrnillan, 1978, Chapter 3. , DEOBOLD B. Understanding Educational Research: An VAN DALEN Introduction, 4th ed. New York: McGraw-Hill, 1979, Chapters 1 and 2. WATANABE, RUTH T. Introduction to Music Research, Englewood Cliffs, N.J .: Prentice-Hall, 1967, Chapter 1. WHITNEY,FREDERICK L. The Elements of Research, 3rd ed. Englewood Cliffs, N.J .: Prentice-Hall, 1950, Chapters 1 and 2. Research Methods in Education: An Introduction, WIERSMA, WILLIAM. 2nd ed. Itasca, 111.: F. E. Peacock, 1975, Chapter 1.

The Setting for Research

29

WILLIAMSON, JOHNB . , DAVIDA. KARP,and JOHNR. DALPHIN.The Research Craft: An Introduction to Social Science Methods. Boston: Little, Brown, 1977, Chapter 1. WISE,JOHNE., ROBERT B. NORDBERG, and DONALD J. REITZ,Methods of Research in Education, Boston: D. C. Heath, 1967, Chapter 1.

Formulating the Problem

I

2 . Formulating the Problem Formulating the problem is unquestionably the most important initial step in research. As noted in the previous chapter, a problem that uses reflective or critical thinking as evidenced in the scientific method is much easier to bring to a satisfactory solution than one that does not. The most profound as well as the least significant problems may prove to be disconcerting to music educators, but if ways are found to resolve this dilemma, their teaching will be more effective. The realization that many problems still remain for the researcher in music education has both good and bad aspects. Most impelling, perhaps, is the negative connotation, because it suggests that conditions, however they

31

may be defined, could be better for those concerned if solutions were forthcorning. On the positive side, graduate students frequently may become unduly apprehensive and consequently apply themselves diligently lest someone else "preempt " their topic and complete the research first. Such misgivings usually are needless because, as most research studies will bear out, the Same problem could be approached differently by two individuals. In addition, when one problem seems to be solved others appear that were not apparent previously . Even though research activity in music education recently seems to have reached an all-time high, and although solutions to many previously disturbing problems have been obtained by an ever-increasing number of investigators, there are other perplexities in which the explanations obtained have been either inconclusive or unconvincing. Perhaps it is one of the ironies of research that indecisive answers frequently result when a researcher fails to give enough thought to the organization and formulation of the problem. In the desire to begin collecting research data as soon as possible, investigators have begun the actual research process before comprehending its significance. An unorganized accumulation of data can be the result of this kind of impulsive action. This writer, in an article in a recent professional journal, called attention to this dilemma.' More than one adviser has had to remand students back to the reflective or critical thinking stage because they had begun to collect disparate data with no reason other than it was interesting to do so. Research data by themselves are virtually worthless unless there is a rationale for using and interpreting them. Commenting on the importance of circumspect attention to the problem, Borg and Ga11 relate that the distinction between an outstanding project and one that is unscholarly does not rest on the "amount of work required to carry it out, but the arnount of thought that the student applies in the selection and definition of the p r ~ b l e m . " ~

Research in Music Education 32

Sources of Information In the formulation of a problem, application of the time-wom "who, when, where, what, why, how " can be most beneficial to an investigator. Graduate students, in a sincere and diligent endeavor to ferret out unique information, often overlook some of the most obvious sources available to them. Some of the most significant ideas for research projects in music education may come from the following, not necessarily listed in order of importance: (1) textbooks relating to various aspects of music; (2) professional and privately printed journals for music educators; (3) lists of graduate theses, dissertations, and projects; (4) reports of subsidized research; (5) graduate seminars and classes; (6) conferences with advisers and professors; (7) discussions with fellow graduate students; (8) difficulties pinpointed in the classroom; (9) speeches and lecture-demonstrations at national, regional, and state meetings of music educators; and (10) visits to libraries, archives, museums, etc.

Textbooks are an unusually good source of potential research topics because quite often readers will find statements with which they disagree. Even more important, a perusal of textbooks places students in a better position to know what is happening in their own field and thus be more knowledgeable about what areas remain as fertile fields for research. Although the concepts and ideas in a textbook may be out-

Forrnulating the Problem

33

moded by the time it is actually printed, investigators can stay up-to-date on the latest trends and ideas in their field by reading periodicals and journals, such as the Journal of Research in Music Education, Music Educators Journal, Bulletin of the Council of Research in Music Education, and College Music Symposium. State journals published by music educators' associations may be helpful, as well as such privately oriented publications as The Instrumentalist, Woodwind-Brass und Percussion, and many periodicals in other disciplines that contain articles, book reviews, and other items relating to music education. Some of these, like the Music Educators Journal, contain accounts of successful techniques teachers have used to surmount some of the perplexities they have faced. Others, such as The Instrumentalist, often include provocative articles that provide possibilities for research topics. One of the most effective ways for students to learn what has been done in the field of music education is to examine lists of theses, dissertations, and projects that are either in Progress or have been completed. After scrutinizing these, researchers frequently See neglected areas in which they feel confident to pursue research. On the other hand, they may discover that their "redhot" idea already has been competently investigated by others. This, of Course, does not rule out the prerogative of continuing research in the Same area, but students should certainly give top priority to a very careful examination of studies already completed so as to make an early decision whether to implement or abandon the idea. Bibliographical and other sources will be discussed in Chapter 4. ~ l t h o u g hfederal, state, and private subsidies for research in music education are not as abundant as they were a few years ago, they did Open up a whole realm of possibilities for investigators, especially at the postgraduate level. Funding is now largely from foundations, and is oriented more toward "related arts," which include art, dance, music, and theater, rather than music alone.

Research in Music Education

34

One example is funding of postdoctoral research by the American Council of Leamed S ~ c i e t i e sLists . ~ of projects become available at intervals, and reference will be made to them in subsequent chapters. Since many of these reports contain suggestions for research needed to follow up the completed studies, they usually provide abundant information for investigators seeking a topic. The interplay between professors and students in graduate seminars and classes often proves to be a source of subjects for research. Possibilities of which students were unaware or had only considered lightly frequently manifest themselves in graduate seminars. This is especially true when emphasis in the class is upon reflective thinking which rnay result in students' clarifying a topic that they already have in mind or in compelling them to begin an analytical search for one. It has been said however, that all people have problems, although they rnay not always be aware of them. The process of interaction with an adviser may help a Student delineate some of them. In the classroom the instructor's comments or remarks often point out areas or problems in need of attention. Graduate advisers should do more than just affix a signature to students' Course prograrn cards, nor do true "advisers" arbitrarily assign topics to their advisees. The students must feel that the subject they eventually select and then implement is their own personal property. How can they spontaneously generate a "personal feeling" about something that has been forced upon them? Advisers should manifest a sense of obligation by making suggestions and assisting each advisee in the formulation and development of the research topic. Quite naturally, professors, because of their experience in the fiel4 rnay have definite ideas about needed research and they should not hesitate to express them. They must, however, be careful to advise, not mandate. Informal discussion among graduate students often results in tangible suggestions for research topics. An especially perceptive individual rnay pinpoint the need for research in an area in which he himself does not feel capable. Yet another qualified Person,

Formulating the Problem

35

who had never regarded this as a possibility for herself or rnay have rejected it as being impractical, rnay seize upon the idea. As a result of discussions with some of her peers she rnay redirect her thinking to a previously discarded subject. On the surface, it would appear that a cataloging of problems encountered in teaching might be a profitable source of research subjects. A more careful study of the prospective topics, however, will very likely reveal that many indeed are in need of study, but the impracticability of defining them clearly so they can be studied in an objective manner tends to exclude some. The dilemmas faced by one researcher rnay serve as natural points of departure for the identification of potential research topics by others. Remarks made by speakers, either informally or in prepared lectures and demonstrations, sometimes supply ideas for research topics. Points of disagreement especially can serve as possibilities for research. Graduate music-education students who are seeking a topic and Want to keep their "ears to the ground" should attend meetings of professional associations, such as the Music Educators National Conference, the Music Teachers National Association, and the College Music Society. Although they reasonably would expect to glean more ideas from meetings in their primary field of interest, they should not overlook the possibility of obtaining stimulation in areas that ostensibly have little interest for them. Visits to libraries, archives, and museums can be especially beneficial, provided a researcher is interested in more than merely becoming farniliar with the holdings. For exarnple, an examination of the writings and effects of certain individuals might reveal the need for in-depth studies about them. A musician could approach the matter from the standpoint of the contributions these individuals have made to music, especially if their claim to fame rests on accomplishments other than in music. For instance, the involvement in and contributions to music of Benjamin Franklin or Thomas Jefferson might prove to be intriguing to some investigator after examining personal effects and writings of these

Research in Music Education

36

esteemed patriots at the Free Library of Philadelphia and at Monticello, Virginia, respectively. Eager, dedicated, and inquisitive researchers should experience little difficulty in finding subjects of particular interest to them. The possibilities are legion, and a constantly changing society and new methods and resources are continually opening up new ones. Researchers can only hope that they can find topics that will be acceptable to their advisers and that can be implemented without undue hindrances.

Vulnerability of Music-Education Research Music-education research often has been vulnerable to attacks not only from leaders in other disciplines but also from members of the profession. In many instances the claims have been justified; in others, not so. Music basically is a skill, and this, no doubt, makes it difficult for a researcher to state in objective terms what actually takes place while the skill is being performed. Conversely, cognitive concepts are sometimes difficult to translate into objective skills. To Pose an obvious question, how can one objectively measure the effectiveness of a teacher of violin in a private studio as against his performance in a classroom with a group of violinists? To be Sure, certain subjective judgments can be made, based on an evaluation of the performance of his students, but this is an indirect type of measurement for a researcher who might have as her topic "A Study to Compare the Effectiveness of a Violin Teacher in Private Studio Teaching as Contrasted to His Classroom Violin Teaching. " This indirect measurement results in a determination of pupil performance rather than the direct evaluation of the effectiveness of the teacher himself. It may be validly argued, however, that pupil

Formulating the Problem

37

to some extent is governed by teacher effectiveness. The distinction, then, between the subjective measurement of a teacher's skill and the objective measurement of pupil performance is a rather tenuous one. Does the latter result from the former, or not? Researchers in music education and other areas have undertaken several studies similar to this suggested topic, often presenting results that are not too convincing. Realizing that research in music education generally has been suffering from what properly might be termed a "research malnutntion syndrome, " the Music Educators National Conference appointed a committee on Graduate Studies in Music Education a few years ago to evaluate research studies in music education that had appeared up to that time. The committee's incisive and candid report was a landmark contribution to knowledge about research in music education. One or more of the ten deficiencies noted then are still evident in many of the projects being submitted "in partial fulfillment" of degree requirements. In paraphrased version, these shortcomings are: (1) a topic that is relatively insignificant; (2) the tabulation of a large amount of miscellaneous data without the conscientious and serious projection of logical conclusions; (3) a project that requires little or no background in music and conceivably could have been completed by a musically illiterate scholar; (4) a study containing errors that might have been eliminated by recourse to original source material; (5) presentation of erroneous conclusions because the investigator did not adequately prepare, collect, and analyze data;

Research in Music Education

38

(6) obvious shortcomings in musical taste and understanding; (7) an inability of the investigator to write concisely and clearly; (8) a failure to reveal intellectual curiosity, which suggests a greater concern in meeting formal Course requirements than in becoming involved reflectively in the quest for knowledge; (9) a lack of understanding regarding the practical problems and potentialities of rnusic education; and (10) noteworthy indications of inadequate preparation and understanding of proper research t e ~ h n i q u e s . ~

Although some studies that seemingly are not significant are still being undertaken, a heartening increase has been observed in research that is not only organized and conducted well, but is focused on a subject that could have considerable import for the profession. Such topics as "A Study to Compare the Effectiveness of the Dry Lip as Opposed to the Wet Lip Approach to Trumpet Playing" are being replaced by those with such titles as "A Study to Determine the Effectiveness of Teaching Beginning Trumpet Students with Computer-Assisted Instruction. " Each one has its merits, no doubt, but the former is concerned with a concept that is highly subjective and does not even arouse much curiosity among brass players today, whereas the latter is related to an idea that is receiving increasing attention on the part of music educators, partially because this device offers certain objectivity in the evaluation of teaching procedures. The tabulation of large amounts of miscellaneous data without appropriate interpretation and projection of logical conclusions is still an area that needs considerable attention. Correctly formulat-

Formulating the Problem

39

ing and implernenting hypotheses that are logical and testable can diminish weaknesses of this kind. False hypotheses delude both researcher and research consumer. Interpretation of these data also requires a researcher who possesses both a perceptive and a creative mind. Research that requires little or no background in rnusic does not seem to be as prevalent in music education today as it once was. It would appear that those charged with the responsibility of directing music-education research studies not only are competent musically, but also, equally important, have an understanding of research techniques and procedures. Persons from other disciplines who wish to pursue a study in the area of music evidently are being discouraged if they do not have an adequate musical background, or they are being required to strengthen their knowledge of rnusic before proceeding with the research. Studies containing errors due to the failure of the investigator to ferret out original source material continue to be more abundant than they should be. With the pressure incumbent upon receiving a degree as expeditiously as possible, such is to be expected. It is unfortunate that some graduate students, often not by choice, rnust be rnore concerned with meeting certain deadlines than with the quality of the research they report. With rising costs of education and of needed supplies, the researcher often finds the proverbial vise closing tighter and tighter. It is hoped that larger amounts of financial Support will become available from govemment agencies and universities so the anxiety and pressure to complete a graduate degree will not be a deterrent to high-quality research. The lessened pressure to meet certain deadlines, because of financial exigencies, will make it easier, also, for research advisers to make more effective appraisals of students' work. In a more relaxed atmosphere careless errors may be recognized and rectified more easily. Closely related to the previous research deficiency is the proffering of erroneous conclusions because the researcher did not

Research in Music Education

40

prepare, collect, and analyze data adequately. Many studies still evidence weaknesses in this area. It is easy to rationalize that this inadequacy also is due largely to the pressure on students to receive their degrees as quickly as possible. In several of the studies in music education that display such shortcomings the explanation lies elsewhere. Sometimes the fault appears to be with the candidates' sponsoring committees, but more often it is the result of their own intellectual shortcomings. Projects organized according to the scientific method and implemented accordingly will rarely exhibit the weaknesses just mentioned. In this connection, Van Dalen appropriately remarks that proper organization of a research proposal is really the most significant step in research, and implementation is "largely a mechanical process which requires more persistence than profundity. " 5 Studies marked by shortcomings in musical taste and understanding are not as prevalent as they once were. In common with the third deficiency, which was discussed abov&, it appears that students with adequate background who may have proposed to engage in some type of nebulous research bearing a general relationship to music are being urged either to pursue their topics in greater depth or to abandon them entirely. The inability to express oneself concisely and clearly is a perpetual problem for many music-education students, one for which satisfactory answers may never be found. Presumably, graduate students, prior to their matriculation for an advanced degree, have had adequate courses that prepare them to express themselves intelligently through writing. It does seem ironic that students who communicate very expressively through the medium of performance sometimes cannot communicate effectively in their native language. (This suggests a topic for research: is there a correlation between performance skills in music and verbal skills?) Many colleges and universities find it necessary to offer special courses in remedial writing for graduate students. But does a remedial writing course of one or two Semesters enable a student

Formulating the Problem

41

to overcome a deficiency that evidently has persisted at least since high school? It is rather difficult to imagine that a one- or two-semester remedial course in music theory, for instance, could enable a student to compensate for a minimum of two to three years of college or university theory instruction that the student either has not had or that was taught so ineffectively as to be meaningless. Some serious resemations exist about the advisability of permitting students who are unable to express themselves adequately in the English language to matriculate for a graduate degree. Might not this serious deficiency indicate the presence of others that would make the successful completion of graduate study extremely improbable? On the other hand, assuming that the graduate students can express themselves well in the English language, the requirement of a master's thesis makes excellent sense, because more and more master 's-degree recipients are continuing on to the doctorate. Experience with the formulation and development of a master's research topic is excellent training for the future doctoral candidate. The failure to reveal intellectual curiosity is a weakness of most serious consequence, one that still prevails in many graduate music-research projects. It seems rather apparent that there is a definite relationship between this defect and some of the others previously discussed, notably the second, fifth, sixth, and seventh. It is a matter of record that in most states it is mandatory for a teacher in the public schools to earn a master's degree or its equivalent in order to qualify fc~rpermanent certification and subsequent advances in salary szhedule. Whether or not this "pressure" has resulted in a lowering of graduate standards is a debatable question that is beyond tiie scope of this book. It is an issue that cannot be resolved by subterfuge. A thorough and impartial reappraisal of the entire graduate program in music education, especially at the master's level, is implicit. One can only wonder how realistic a grading system is in graduate schools that recognize only "Au and "B" as passing grades. Furthermore, some schools that formerly permitted only

Research in Music Education

42

"AM or "B" grades now have further muddied the waters by adopting a grading System that permits "A, " "A-," "B+," "B, " and "B - . " In effect, this dilution may actually be more representative of the student's achievement than the Standard "A" or "B. " Do "average" students suddenly become "good" or "excellent" when they reach graduate school? Proponents of this system contend that "average" students on the undergraduate level are not admitted to graduate study, so therefore only those who are "good" or "excellent" remain. Unfortunately, not all graduate schools share this concept; many of them admit students according to other criteria, even when the students do not possess the requisite "B" average for admission. It seems likely that there will be more rather than less relaxing of the "B" requirement for graduate admission in the future, coupled with increased emphasis on other factors that may be more truly indicative of a student's probable success in graduate study. It is possible that some of these factors, such as natural curiosity or motivation, which may not always be apparent in students' grades, can be used to determine whether to adrnit students to graduate study. These individuals may become so engrossed in their subjects that they will experience less difficulty in pursuing research to a logical and successful conclusion than will others who do not have the Same motivation. The performance of veterans who enter college or university with previous unspectacular records is a case in point. Some students who have neither the background nor understanding of the discipline enroll in graduate music-education programs because it seems expedient to do so. Prospective doctoral candidates often remark candidly that it is "easier to get a doctoral degree in music education than in other areas of music. " If this indictment is true, it is time for a review of doctoral programs in music education to determine areas that need strengthening. A few years ago, when more teachers were needed to meet the demands of an exploding population, students entered graduate study in music education with an undergraduate background in the liberal

Formulating the Problem

43

arts. Many did not have general and specific methodology courses, psychological principles of education, skill in the performing areas of music, or Student teaching. To engender an understanding of the philosophies, principles, and problems of rnusic education is not easily accomplished in two or three Semesters of graduate instruction. 1s it not reasonable to assume that those who have taught in the public schools are in a better position to cope with practical problems and potentialities in the field when they enroll in a graduate program of music education than are those who lack such experience? A new cloud is appearing on the horizon. With the decreasing birthrate and the closing of both elementary and secondary schools due to declining enrollment, colleges and universities will themselves begin to experience a severe decline in enrollrnent around the middle 1980s or the early 1990s. This means that cornpetition for students will become even rnore intense, resulting in a lowering of Standards to admit unqualified students to graduate music-education programs. Fortunately, the increase in graduate courses in music education has been accompanied by a recognition of the responsibility on the part of institutions to offer courses that devote some attention to research techniques. This factor, perhaps more than any other, has resulted in a gradual increase in both the quantity and quality of research studies in music education. There seems to be an unanimous agreement among researchers that good research usually does not just happen. Addressing themselves to this point, Barzun and Graff affirm that "reading, writing, und thinking are the three activities of research. "6

Initial Considerations in the Selection of a Research Problem Most textbooks on research methodology contain detailed and helpful suggestions of a general nature pertaining to the selection of a problem. One of the more concise accounts is that by Cham-

Research in Music Education 44 bers in the journal of a professional education society.' Some of the important considerations listed therein are concerned with the cordial relationships that should exist between students and advisers as the former seek to select problems for research. The proposals of Chambers, in paraphrased version, indicate that potential researchers should: (1) avoid asking their advisers for "assigned" topics, but rather seek those that are in accord with their own interests and initiative; (2) elect subjects that are in harmony with their interests and backgrounds instead of those that are suited to the "predilections" of their advisers; (3) manifest erudition by not ,expecting their advisers to serve as "intellectual nursemaids ' '; (4) define their problems clearly; (5) become familiar with literature in their field so as to find out what has or has not been done; (6) determine what methods, techniques, or instruments will be needed; and (7) find out whether field trips or visits to museums, libraries, private archives, and other repositories of information are necessary. After continuing with suggestions for student-adviser conferences and for writing up the study, Chambers concludes by stating: "Research is not necessarily as complex, difficult, mysterious, or esoteric as a pedantic attitude can make it seem. In common with all things that are really great, it is essentially simple in concept.

Formulating the Problem

45

It has been comprehensively and simply defined in eight words as 'the orderly treatment of data to answer questions. ' "8 It should be noted that topics that are "assigned" by advisers have validity for investigators only if they are interested in the subject und can involve themselves wholeheartedly in it. By analogy, there are too many exercises in music theory that result in a mediocre or mechanical sound because the people writing them have not been given the freedom to express themselves in a manner that is meaningful and significant to them. This, obviously, is not to imply that all principles and rules should be abrogated. They are certainly needed as guidelines, but students need to be encouraged to express themselves in a manner that is in accord with their own initiative and creativity. All too prevalent are research projects in a "series," usually at the master's level, which are "assigned" by advisers to certain of their students. A group of studies of this type might center around a general title, such as a survey of elementary private method instruction books for specific instruments, with students examining and comparing materials for each instrument of the band and orchestra. Such projects if well organized and implemented can be very beneficial to students in instrumental-techniques classes who are unfamiliar with these publications. It is in this Spirit, no doubt, that advisers make such assignments. Koefod takes a less optiniistic viewpoint: "Too many students do nothing but plead for assignment of projects by faculty supervisors, apparently having been somehow conditioned by the 'system' to do t h i ~ . " ~ Whether the "assignments" are voluntary or involuntary, the results will usually be the Same unless the students are interested and completely immersed in the subject. The process of reflective thinking discussed earlier in this chapter will be more effective when the choice of a topic is the result of the students' initiative, since the decision undoubtedly will be one of the most important they will make in their educational careers, and thus should be predicated on something in which they are intensely interested.

Research in Music Education

46

Chambers's reference to advisers serving as "intellectual nursemaids" is both amusing and tragic. The implication is that students will merely put the "flesh" on "skeletons" constructed by their advisers. Advisers can no more do students' reflective thinking for them than they can take their examinations. Advisers should direct attention to questionable Patterns of thinking and organization of materials. Their suggestions should be practical, relevant, and within the framework of the research proposal. Advisers' remarks are suggestions, not mandates, with full responsibility for accepting or rejecting them resting with the students, upon whom, also, the onus eventually falls for defending their research. Stating the problem clearly has been discussed earlier in this chapter and needs no additional elaboration here; we need only reiterate its importance. It is incumbent on researchers, as already indicated, to familiarize themselves with the literature in th'eir field. Many advisers continue to be distressed by the inordinately large number of students who come to them for advice regarding a "red-hot" idea before they have examined the literature to determine whether the proposal is feasible. Such a Course would eliminate many projected topics, leaving the students free to devote their energies, and those of their advisers, to topics within the realm of possibility. It is conceivable that two individuals may be proceeding with the Same topic unbeknown to one another. Although the likelihood of this happening is slim, Good and Scates give an account of two music-education research studies conducted on the Same national organization, although they were not done at the Same time. The projects were completed in different sections of the country even though the institutions were only three-hundred miles apart, but the initial study, which was finished five years before the second, could easily have been located by the second investigator in Doctoral Dissertations Accepted by American Universities. lo

Formulating the Problem 47 After their topics have been tentatively accepted, with the assistance of their advisers, investigators should carefully consider and make a projection of the methods, techniques, equipment, and instruments they will use to bring their studies to successful fruition. This might involve the use of certain standardized tests, instniments of the investigators' own devising, questionnaires, and Special equipment, such as teaching machines, Computers, or other devices. Need for these items will be determined partially by the methods or techniques chosen by the researcher. For example, an investigation based on the experimetal method normally would not utilize the questionnaire technique, whereas one based on a descriptive study would not often use the concept of the Single variable (to be discussed in a later chapter). An investigator involved with historical, and frequently a philosophical or aesthetic, study will need information that may be obtained from various repositories (libraries, museums, archives, historical societies, private collections, etc.). Since this will be treated in greater detail in Chapter 5, mention will be made here only in passing. The researcher will need to know where specific items are located. If visits are involved, the investigator must plan them according to such considerations as financial resources, time available, and hours the repository is open. Frequently it is next to impossible to conduct systematic and exhaustive historical research without this kind of planning. Field studies, in the form of visits to other schools or colleges and universities, often are necessary in a descriptive study when comparisons are being made between curricular offerings or programs of music education. To determine why a siring program may be successful in one school System but not in others can hardly be done properly by means of a questionnaire only. Interviews, personal visits, and a questionnaire might be needed to garner enough information to make valid Statements when comparing schools. Reference will be made to these techniques in the chapter on descriptive research.

Research in Music Education 48

Steps in the Formulation of a Research Problem Investigators who have a general subject in mind must proceed through several steps before they reach a point at which their research topic is well defined, practical, and capable of implementation. Almost every textbook offers excellent suggestions relating to educational research. For instance, Rummel states that "personal interest, personal capabilities, value of the topic, and availability of data" are the important principles to consider in the formulation of a research problem.l1 Almack indicates slightly different precepts, including what already is known and what needs to be leamed, inconsistencies evident, and implementing suggestions received from various s o u r c e ~ . ~ ~ Researchers may ask themselves certain questions about the problem they have under consideration. The answers to these queries will go a long way toward determining whether they should proceed with the plan, modify it, or discard it entirely. For example, Hillway poses five questions, of which the first four should be answered in the affirmative and the last one in the negative, if a researcher is to have a sound base for the formulation of a problem. He asks, "1s the proposed topic interesting?" "1s it novel?" "1s it requisite to advance knowledge?" "1s it practicable? " and "Has it already been investigated? "I3 By posing and answering general questions similar to those of Hillway music educators may then proceed to more specific items in an attempt to delineate the problem even more clearly. They might, for purposes of illustration, begin with a topic like "A History of Music in the United States. " At first glance, the subject may appear to be a good one for a graduate student, especially at the doctoral level. Before proceeding, however, a researcher would need to ask such questions as: Will the study Cover all phases of music (instrumental, vocal, Opera, symphony, chamber music, music education, etc.)? Will it be concerned with

Formulating the Problem

49

teaching, or both? What years will serve as the limits of the study? What kind of information will I seek? Where may I best obtain the desired information? How will I interpret the phrase "Music in the United States"? Will the study involve only native-bom American musicians? Will it include the role of music publishers? Will it analyze the compositions mentioned or only identify them without comment? Will the study include artists and performing groups from other countries? What distinction, if any, will it make between amateur and professional groups? Will I collect and catalog manuscripts or facsimile copies of works by American composers? What attention will I give to reviews of performances presented by artists in various cities? and How will I determine what gaps still remain in American musical history? Almost immediately it should be apparent that "A History of Music in the United States" is a subject that consists of many components, each of which might be a topic worthy of investigation. This matter of delimitation is very important for successful research, yet investigators in music education frequently do not accord it enough attention. While there are certainly enough topics to go around, graduate students are unduly apprehensive about the possibility that all areas in which they are interested soon will be fully explored. Actually, as research studies are completed they commonly suggest new areas needing investigation. It has already been emphasized that a study that is well organized and implemented in depth can make a much more significant contribution to human knowledge than one that is so broad that the generalizations it reaches result only in superficial information. The foregoing fourteen questions are suggestive of those that researchers must keep in mind as they attempt to focus their attention on the hypothetical research problem just noted. The feasibility of any topic may be determined only after similar questions have been asked. In other instances the impracticality of a plan as originally conceived will be apparent earlier. After

Research in Music Education 50 revisions researchers will repeat the Same procedures until they conhave developed a satisfactory proposal. This process-f tinually returning to the begiming and repeating the steps until satisfactory results are obtained-is reminiscent of the looping technique used in certain phases of programming for the digital computer, where a program is repeated in successive steps until the desired information is received. In both instances the Same questions may be asked over and over again, but with different objectives in mind. Sequential steps in the actual formulation and implementation of a problem are to: (1) determine an area of need to which answers are not obvious; (2) ascertain whether or not the idea contains more than one basic problem; (3) delimit the subject to a topic that may be solved according to the background and training of the investigator; (4) develop basic assumptions andlor hypotheses; (5) locate existing information relating to the topic; ' (6) ascertain what instruments, tools, or equipment will be needed and whether such items are readily available or may be devised expeditiously; (7) postulate tentative conclusions; (8) implement the plan and accumulate preliminary data; (9) reconstitute and revise the research plan in view of inaccuracies appearing as a result of the preceding steps; (10) accumulate and assimilate additional data;

Formulating the Problem

51

(1 1) interpret the data; (12) draw up conclusions, recommendations, and suggestions for additional research; (13) obtain an evaluation of the research from peers and associates; and (14) disseminate results to the general public.

Determining an area of need for which answers are not obvious may seem to be redundant, but it is only through this type of intellectual inquiry that investigators really can decide whether they have the basis for a research topic. Answers to questions frequently are available in many sources, although the researchers may be unaware of them. However, the need for a detailed study of an area of concem may not be as critical as the researchers had originally believed. An instrumental-music teacher, for instance, who is interested in obtaining background material on Berlioz to serve as motivation for introducing a high-school orchestra to the composer's King Lear Overture hardly needs to go beyond Jullien's biography or Boult's Berlioz's Life as Written by Himself in His Letters und Memoirs, to cite two important sources. The attempt to locate some new or little-known facts of the composer's life, although it could prove interesting, would hardly be worth the time and effort necessary to prepare for this orchestra rehearsal and might in addition prove to be fruitless. If, on the other hand, a researcher was concerned with one specific aspect of Berlioz's life then the research, although time consuming, might result in a noteworthy contribution, such as Barzun's publication of some previously unknown Berlioz letters.14 Experienced researchers like Barzun, of Course, usually find it easier to obtain data because they know what they are seeking, whereas inexperienced researchers often do not. Hillway is among the many writers who Stress the importance of investigators' approaching research with a good idea of what they hope to find 0~t.l~

Research in Music Education

52

After careful scrutiny an initial idea may be subdivided into several problems, each of which might constitute a topic in itself. "A History of Music in the United States" obviously needs to be divided into several topics before it can be considered feasible and practical. It has already been observed that it is very difficult to apply the techniques of sound research to a topic that is too broad in scope. Such a subject disperses a researcher's energies in several directions. Figure 1 serves as an illustration of this process of delineation. As a result of successive delimitations, "The Development of Municipal Professional Symphony Orchestras in the State of Florida from Earliest Times to 1965 " is a subject for research that a student with appropriate background training and motivation could conceivably pursue. The final title not only confines the study to certain specific objectives but also enables the investigator to proceed with research once the modus operandi is established. The term "professional," as used in this illustration, might be defined differently for Florida than it would be for New York City, but the study would be just as valid in either instance if the research were properly constituted and implemented. After determining the feasibility of the topic, at least tentatively, students must then consider certain steps, when they are needed in the plan of implementation; namely, basic assumptions andlor hypotheses, terms that will be discussed in greater detail in the next chapter. It should be noted, however, that assumptions and hypotheses serve as the core of a study. A basic assumption refers to what an investigator assumes to be true, and thus does not need to be verified. A basic hypothesis is an educated guess that will need to be confirmed by experimentation, rationalism, or other means. After basic assumptions and hypotheses have been formulated students must determine the kind of research plan or methodology best suited to the nature of their research problem. It is at this point that researchers in music education will decide whether

Formulating the Problem

53

Figure I Delimitation of a Research Topic

TENTATIVE TITLE

How DELIMITEDFROM PREVIOUS LISTING

A History of Music in the

United States A History of Instrumental Delimited to instrumental Music in the United music States A History of Instrumental Further delimited to one Music in the Southem region of the United United States Stares A History of Instrumental Additional chronological Music in the Southern delimitation United States from Earliest Times to 1965 The Development of Slight change of title and Municipal Symphony further delimitation to Orchestras in the specific form of instruSouthern United States mental ensemble from Earliest Times to 1965 The Development of Additional delimitation to Municipal Professional exclude amateur groups Symphony Orchestras but to include a specific in the State of Florida state of the United States from Earliest Times to 1965

Research in Music Education

54

their study will be essentially historical, experimental, descriptive, philosophical, aesthetic, or a combination of these. The ultimate determination is largely predicated on the previous steps. Knowledge of relevant information is an important next step. Students who, searching through various resources, both published and unpublished, find that their chosen topic has already been adequately investigated should either recast the problem or abandon it altogether. A paucity of information, of Course, suggests fertile areas for investigation that eventually might result in meaningful contributions to human knowledge. Moreover, a wealth of material might enable investigators to concentrate on a more specialized phase of the problem. Their methodology will be governed to some extent by the kind of data they seek. An experimental project, which would have as its end result the formulation of conclusions to controlled observations, would necessitate the quest for a different type of data than would a historical study, which might rely heavily on original manuscripts or documents found in archives. In either instance the information sought must be pertinent to the topic at hand. Experimental and descriptive studies, in particular, frequently make use of tests, statistics, special equipment, or questionnaires. Sometimes it is necessary for the researcher to devise and validate tests or special equipment with the specific purposes of the research in mind; in others standard items rnay be used. Investigators devise questionnaires, almost without exception, to fit the needs of a particular study, usually descriptive. Unlike a test, which rnay be validated or standardized, a questionnaire rarely can be used by more than one researcher unless another is investigating the Same problem under comparable conditions. Researchers rnay use standardized tests to supplement those they have devised themselves. When tests, equipment, or instruments are not available, investigators must either develop them themselves, adapt existing tests for purposes of their study, or revise their methodology to use those in existence.

Formulating the Problem

55

The postulation of tentative conclusions helps investigators to anticipate the results of their research. To paraphrase a cliche, if researchers do not know what they are seeking, how will they know when they solve the problem? Tentative conclusions essentially are postulated from the basic hypotheses. It is through implementation of the plan and accumulation of preliminary data that researchers can really test their hypotheses. They are then in a position to know whether and to what extent their research plan needs to be revised. After necessary revisions, additional data are accumulated and assimilated. Mere accumulation of data, however, is not enough. They will need to be interpreted in light of objectii-)esof the study. It rnay be of little practical value, for instance, to find out that there are two hundred school-owned oboes in the public schools of a certain state. More important is knowledge about whether or not these instruments are in playing condition, how many of them are being used and by whom, and the proficiency level of the students playing them. Moreover, the accumulation of additional data rnay be necessary in order to make logical and valid interpretations. Conclusions, recommendations, und suggestions for additional study are commonly grouped together in the final chapter of a research report. Final conclusions, although they rnay be similar to the tentative conclusions identified earlier, are deduced after the data have been accumulated and interpreted. Recommendations and suggestions for additional research then follow in normal sequence after the conclusions have been drawn. The recommendations usually include areas that, as they reflect on the project, researchers feel might either have been approached differently or pursued in greater depth. On the basis of these, researchers offer suggestions for the benefit of future students. Often the final chapter of a research report is the most vital one to someone searching for a topic, for it is here that the investigator may suggest areas that are in need of study. A critique of the completed research by peers und associates

Research in Music Education

56

can be most helpful. An investigator who wishes to avoid the usual redundancies that characterize the creative efforts of most persons should welcome the opportunity to have an outside obServer examine the completed project objectively and impartially. Because of a natural preoccupation with the project, a researcher can seldom do this. Cntique of research by peers and associates is fairly Standard practice in medical and scientific research, but in music and other areas of education its use has been rather limited. It is for this reason that additional members are usually added to a doctoral candidate's examining committee at the time of the final oral examination, when basically the Student must defend his or her research. A serious weakness in music education has been the disseminarion of research results to the general public. Most completed doctoral projects in music education eventually are listed in Dissertation Abstracts International, but only a very small percentage of the musical public who might be able to use these findings has access to them. Some reports do appear in national and state music educators' or research journals, although usually only as a result of the investigator's initiative in preparing an abstract of the research for publication. Much useful research remains unknown because researchers either do not take the time to write an interesting account of it or feel that their responsibilities have been discharged once the project is completed. This final step in the research process deserves much greater emphasis than it has received up to the present time. If provisions are made to disseminate research results some of the better master's theses in music education could be made available to the musical public. On a more limited scale, Master's Abstracts, containing brief accounts of research by master's candidates, is a companion to Dissertation Abstracts International, but circulation is largely limited to libraries. Other resources will be mentioned in Chapter 4 of this book. It

Formulating the Problem

57

is encouraging, however, to note that many music educators have been distressed by the sparse amount of research information that has been disseminated in a form palatable to the general musical public. Petzold urges that the results of research be shared with t he practicing music teacher, using terminology that is readily understandable, so that any implications the study may have for practices in music education are evident. The teacher is seldom enthusiastic about reading an article presented in the typical research jargon that contains an overwhelming amount of technical information. He wants, and deserves, the Reader's Digest approach which summarizes the essential~in a straightforward manner.16 The finished research project is the result of careful planning and implementation on the part of the investigator. Emphasizing this, Koefod states that the researcher must "establish beyond doubt that he has conceived a meaningful intellectual exercise and carried it through gestation to fruition. '"' Just as an architect's blueprint guides a building contractor, so a research proposal serves as the design that points the way for a researcher. In the next chapter we shall discuss this plan, design, or proposal.

Problems for Revie W and Discussion 1. Give some of the most likely sources of ideas for research projects in music education.

2. What have been some weaknesses of music-education research in the past?

Research in Music Education

58

3. Discuss suggestions of a general nature to be used in the selection of a research problem.

4. The formulation of a research problem is dependent upon what general steps or procedures? What are some of the specific questions investigators might ask themselves as they give consideration to their problem?

5 . What sequential steps are involved in the formulation and implementation of a problem?

6. Take a broad subject area and delimit it to a workable topic. (Use Figure 1 as an example of how a broad subject can be narro wed .)

Supplementary Readings ALMACK, JOHNC. Research und Thesis Writing. Boston: Houghton Mifflin, 1930, Chapters 2 and 4. ARY,DONALD, LUCYC. JACOBS, and ASGHAR RAZAVIEH. Introduction to Research in Education, 2nd ed. New York: Holt, Rinehart and Winston, 1979, Chapter 2. BARZUN, JACQUES, and HENRYF. GRAFF.The Modern Researcher, 3rd ed. New York: Harcourt Brace Jovanovich, 1977, Chapter 2. BEST,JOHNW. Research in Education, 3rd ed. Englewood Cliffs, N. J .: Prentice-Hall, 1977, Chapter 2. BORG,WALTER R., A N D MEREDITH D. GALL.Educational Research: An Introduction, 3rd ed. New York: Longrnan, 1979, Chapters 2 and 3. CHAMBERS, M. M. "Selection, Definition, and Delirnitation of a Doctoral Research Problem, " Phi Delta Kappan, November 1960, 42,2:71-73. COOK,D A V ~RD. , and N. KENNETH LAFLEUR. A Guide to Educational Research, 2nd ed. Boston: Allyn and Bacon, 1975, Chapter 2.

Formulating the Problem

59

DEWEY, JOHN.How We Think. Boston: D. C. Heath 1933, Chapter 7. ENGELHART, MAX D. Methods of Educational Research. Chicago: Rand McNally, 1972, Chapter 3. Fox, DAVIDJ. The Research Process in Education. New York: Holt Rinehart and Winston, 1969, Chapter 2. GAGE,N. L., ed. Handbook of Research on Teaching. Chicago: Rand McNally, 1963, Chapter 2. GAY,L. R. Educational Research: Competencies for Analysis und Application. Columbus, Ohio: Charles E. Merrill, 1976, Chapter 2. GOOD,CARTER V. Essentials of Educational Research. New York: Appleton-Century-Crofts, 1966, Chapter 2. "Graduate Study in Music Education," Journal of Research in Music Education, Fall 1954, 2,2: 157- 170. GRIFFITHS, D A N ~ EE. L Research in Educational Administration: An Appraisal und a Plan. New York: Bureau of Publications, Teachers College, Columbia University, 1959, Chapter 1 . HILLWAY, TYRUS. Introduction to Research, 2nd ed. Boston: Houghton Mifflin, 1964, Chapter 8. HOPKINS, CHARLES D. Educational Research: A Structure for Inquiry. Columbus, Ohio: Charles E. Merrill, 1976, Chapter 4. JONES,RALPHH., ed. Methods und Techniques of Educational Research. Danville, 111.: Interstate Printers and Publishers, 1973, Part 1. KOEFOD,PAULE. The Writing Requirements for Graduate Degrees, Englewood Cliffs, N. J .: Prentice-Hall, 1964, Chapter 1. MASON,EMANUEL J., A N D WILLIAM J. BRAMBLE. Understanding und Conducting Research: Applications in Education und the Behavioral Sciences. New York: McGraw-Hill, 1978, Chapter 3. MOULY, GEORGE J. The Science of Educational Research, 2nd ed. New York: Van Nostrand Reinhold 1970, Chapter 4. PHELPS,ROGERP. "The Doctoral Dissertation: Boon or Bane?" College Music Symposium, Fall 1978, 18,2:82-93. RUMMEL, J. FRANCIS. An Introduction to Research Procedures in Education, 2nd ed. New York: Harper and Row, 1964, Chapter 2. RUNKEL, PHILIPJ., and JOSEPHE. MCGRATH. Research on Human Behavior. New York: Holt, Rinehart and Winston, 1972, Chapter 2. SAX,GILBERT. Empirical Foundations of Educational Research. Englewood Cliffs, N. J.: Prentice-Hall, 1968, Chapter 3.

Research in Music Education

60

TRAVERS, ROBERT M. W. An Introduction to Educational Research, 4th ed. New York: Macmillan, 1978, Chapter 4. ed. Second Handbook of Research on Tesching, Chicago: Rand McNally, 1973, Chapter 1. TUCKMAN, BRUCEW. Conducting Educational Research, 2nd ed. New York: Harcourt Brace Jovanovich, 1978, Chapter 1. VANDALEN, DEOBOLD B. Understanding Educational Research: An Introduction. New York: McGraw-Hill, 1979, Chapter 7. WATANABE, RUTH T. Introduction to Music Research. Englewood Cliffs, N.J .: Prentice-Hall, 1967, Chapter 8. WHITNEY, Frederick L. The Elements of Research, 3rd ed. Englewood Cliffs, N.J.: Prentice-Hall, 1950, Chapter 3. WIERSMA, WILLIAM. Research Methods in Education: An Introduction, 2nd ed. Itasca, 111.: F. E. Peacock, 1975, Chapter 3. WILLIAMSON, JOHN B., DAVID G . KARP,and JOHNR. DALPHIN. The Research Craft: An Introduction to Social Science Methods. Boston: Little, Brown, 1977, Chapter 2.

,-

3. Organizing the Research Proposal The prospectus, which graduate students subrnit for approval before they officially undertake their research, is known by various terms: thesis proposal, thesis outline, research agendum, research design, or dissertation proposal. Regardless of the nomenclature, the purpose is the same-for the investigators to indicate systematically, in writing, the plan whereby they propose to collect, organize, and interpret the data of their research. It is imperative that researchers give careful consideration to the preparation of their research proposal. Griffiths points out that careful researchers are never satisfied with things the way they are, are constantly seeking to obtain clearer explanations of obscure areas in their field, and are concerned with developing a taxonomy that

Research in Music Education

62

pulls together knowledge in their field.' In a recent publication this writer stressed the importance of a good research proposal that is worthy of sympathetic guidance from the Dissertation Committee, also known as a Sponsoring Committee.* After official acceptance by a committee or an individual especially designated to evaluate such proposals, the researchers ' plan functions as the outline or guide as they proceed with their research. The approved design also serves as the instrument they will use to obtain additional assistance and guidance from an adviser or Sponsoring C ~ m r n i t t e e Best . ~ considers the research design analogous to the architect's blueprint, which must be prepared before construction can begin on a b ~ i l d i n g . ~ The research proposal, even after it has been officially accepted, should not be considered as immutable, for changes rnay prove to be advisable as the actual collection of data proceeds. These modifications rnay be minor, such as-a slight change in title, or major, resulting in extensive alterations. As an example, subsequent to the acceptance of "An Experimental Study to Determine the Effectiveness of Teaching Begiming Violin Students by Means of a Teaching Machine, " it might be deemed advisable to change the title to "An Experimental Study to Determine the Effectiveness of Teaching Beginning Class Violin Students by Means of a Teaching Machine. " In the original title no suggestion is given that the teaching will be for a class situation, even though it rnay be so delineated in the actual body of the proposal itself. The subsequent revision clearly indicates this clarification. A research title, about which more will be said later in this chapter, needs to be a succinct and direct Statement of the exact nature of the proposed study. Other changes rnay assume greater import even to the point, in extreme instances, of a radical change in methodology. For example, in neither of the titles just proposed relating to violin teaching is the methodology explicit. It is implicit, however, to the extent that some kind of comparison is suggested. It is in the

Organizing the Research Proposal

63

actual plan that actual methodological changes rnay take place. These in turn rnay suggest some modification of the title. An illustration: an investigator, having determined that it is not feasible to conduct an experimental study to compare the effectiveness of class violin teaching by means of a teaching machine, instead decides to use two randomly selected violin classes with what Campbell and Stanley call the Pretest-Posttest Control Group Design and compare them e~perimentally.~ More will be said about this technique in Chapter 6. The specific experimental techniques used in this instance would involve giving a pretest to both groups of violin students, then introducing an "X" factor or independent variable to one group, which could be prestudy of the notes prior to playing them. The other group would have no prestudy prior to changes. A posttest would then follow to meaSure any changes in reading skill that rnay have taken place as a result of manipulating the independent variable. Emphasizing the importance of the research design, Asher says: "The most important part of research methods is research design. Knowledge of research design is the methodological base upon which the profession rests. . . . Research design Opens many doors to knowledge. "6 Although there is no universally accepted format for a research desiga7 the one used in the School of Education, Health, Nursing, and Arts Professions at New York University includes the components usually found in plans in effect elsewhere. This is not to suggest that the format used for research designs at other institutions is less significant than the one to be discussed shortly, but simply that the writer considers this to be one of the better plans with which he has worked. The schema of Figure 2 is flexible, so that the kind of research pursued will largely determine which of the components will be utilized. Modifications and changes obviously are in order when and where the occasion demands. The plan that follows in Figure 2 suggests the kinds of information to be included in a research design.

Research in Music Education

64

Organizing the Research Proposal

The Research Design

Figure 2 Format for a Research Design

Many students make the mistake of spending many fruitless hours groping The Title for a title when the focus of the proposed research is not clear. The title, which should evolve after the problem statement has been inductively determined, is a broad but accurate account of the scope of the study stated in clear, conceptual form. This writer has reviewed all too many studies in which the title and statement of the problem represent entirely different entities. This can be frustrating to someone who receives titles from an information-retrieval System and then, after ordering and receiving the documents, learns that the studies have little relevance to the title. Occasionally a general title will not be complete without the addition of a subtitle, as a subsequent example will show. To repeat: students should not decide on a title before the problem statement is clear in their minds. However, for purposes of illustrating the reflective process involved, a statement follows that will be developed into a title, problem statement, subproblems, basic assumptions, and basic hypotheses. The statement:

Title Subproblems Definitions of Terms Delimitations Basic Assumptions Basic Hypotheses Need or Significance of the Study Related Literature Met hodology Bibliography Personal Qualifications Appendix

A research design should be long enough to indicate clearly what the researcher proposes to do and how this is to be carried Out, but not so verbose that it would appear to be the research design and dissertation or thesis combined in one document. (When this writer served as coordinator of the all-school Research Proposals Committee at New York University a few years ago it was not unusual to receive a research design of 120 to 130 pages.) A researcher should be able to include all that is necessary in a research design numbering no more than forty typewritten, double-spaced pages, exclusive of Bibliography, Personal Qualifications, and Appendix material. Redundant material from the research design should be reserved for the thesis or dissertation. A discussion of each of the components of the research design, with examples for consideration, now follows.

65

The nocturne, a character piecefor piano, was introduced by John Field (1782-1837). As applied to a somewhat melancholy or languid style, with an expressive melody over a broken-chord accompaniment, the term "nocturne" has been used by many composers, but perhaps the individual irlho adapted the idea most successfully was Frederic Chopin. (It is interesting to note that both Field und Chopin it8rote eighteen nocturnes.) As a Pianist you Want to examine the development of the nocturne to assist you in performing nocturnes by various Composers.

I

A title that might be derived from the above material is: "The Nocturnes for Piano of John Field and Frederic Chopin." How-

Research in Music Education

66

ever, when an additional subtitle is added the focus of the study becomes clearer, as follows: "The Nocturnes for Piano of John Field and Frederic Chopin: A Stylistic Study to Aid in Their Performance. ' ' Since the statement of the Problem Statement problem should be a clearly worded indication of the focus of the study, it should be somewhat more specific than the title. The problem statement is a cognitive or conceptual statement of what the researcher wants to find out or plans to do. It does not indicate how the information is to be obtained. It is sometimes referred to as the gestalt or organismic representation of a researcher's problem. A problem statement is usually expressed in declarative form. Compare the hypothetical statements that follow with the title just noted: ( I ) The nocturnes of John.Field and Frederic Chopin are the Same stylistically because of their designation as ''nocturnes. " (2) The nocturnes for piano of John Field are comparable musically to those of Frederic Chopin. (3) The purpose of this study is to compare stylistically the eighteen nocturnes for piano of John Field with those of Frederic Chopin as an aid to their performance. (4) The purpose of this study is to compare stylistically the nocturnes of John Field with those of Frederic Chopin as an aid to their performance. Of these statements, when compared with the original source material and title above, the clearest would be the third one because it encompasses the important items that are necessary to a clear understanding of what the researcher proposes to do. The

Organizing the Research Proposal

67

first statement is not valid because it suggests that all compositions designated as "nocturnes" are the Same. The second statement is not valid because no indication of comparing the composers' nocturnes is given in the original statement. While valid, the fourth statement is not as complete as the third one. The third statement, then, is best because it incorporates more of the information contained in the original statement than any of the others. During the Course of collecting data it may prove desirable to make changes in the title as well as in the problem statement, as noted previously. Whitney voices a generally accepted principle, that an alteration of both research design and title should result when it is evident that the research is proceeding differently than originally p r o p ~ s e d . ~ Sometimes known as "subordinate " Subproblems or ' 'specific problems, " subproblems, either in declarative or interrogatory form, state the atomistic or more precise aspects of the problem statement. In other words, the statement of the problem is broken down into more definite components. Subproblems, however, are not to be confused with subtitle. Subproblems are expressed in cognitive or conceptual terms and like the statement of the problem, indicate what the researcher plans to do or find out. Subproblems are the parts that make up the whole or gestalr. The researcher must find answers to these subproblems in order to resolve the main problem. Subproblems should be proffered in logical order because they indicate the direction the research will take. In many theses or dissertations each subproblem will become a chapter in the final document. Using the hypothetical third statement of the problem noted above, examine these subproblem statements: (1) To compare the musical worth of the piano nocturnes of John Field and Frederic Chopin.

Research in Music Education

68

(2) To trace the background of the nocturnes up to the time of John Field. (3) What System of analysis may be devised to analyze the nocturnes? (4) To determine similarities and differences in the eighteen nocturnes for piano of John Field and Frkdiric Chopin. Of these four Statements, three are in a declarative form, beginning with an infinitive. The third, however, is in an interrogatory form. The two forms of stating a subproblem should not be intermingled; the subproblems should either be all declarative or all interrogatory. The first statement would not make a good subproblem because musical worth is not a concept being compared according to the original statement. The second statement would make a good initial subproblem because background information is always necessary in a study of this type. The third statement, even though expressed here as a question, would be viable if it were in infinitive form because researchers need to adapt some existing approach to musical analysis, or devise their own. Finally, the fourth statement would be a valid concluding subproblem because the intent here is to pul1 together information obtained from all the previous subproblems. A researcher should define any Definition of Terms terms used in the title, problem statement, subproblems, or throughout the study if they are ambiguous or technical. In addition, terminology used to connote a meaning different from that usually accepted should be clarified. This would be true not only for such terms as gebrauchsm~sik,~ which might not be understood by any except the well-read musician, but also for such seemingly well-known terms as "chorus," when the investigator wishes to employ a specific meaning. This is not to suggest that a researcher has license to manipulate a definition to meet the purposes of the study or to use one that is not generally accepted

Organizing the Research Proposal

69

by authorities in the field. The sole purpose of including definition~is to bring about clarity. A well-balanced definitions section contains both direct quotations from the literature and paraphrased terms. Definitions may be c o n c e p t u a l ~ l e a r l yand concisely operational-or indicating what the concept means-orfor instance, indicating what tool or instrument will be used to obtain the desired information. A case in point: a researcher provides a cognitive definition of "creativity" and then states that "creativity " will be operationalized by using an instrument like the Torrance Tests of Creative Thinking.l0 Researchers, moreover, must neither offend their readers by defining terms that are obvious, nor assume that all of their readers are as farniliar and as well versed in their subject as they are. It is this desirable balance for which the researcher must constantly strive. Referring again to the hypothetical problem statement above, study these definitions: nocturne-a musical piece that suggests night visions and dreams, sometimes called nachstuck. nocturne-a piano composition in a melancholy or pensive style, characterized by an expressive melody over a broken-chord accompaniment.ll Of the two definitions, the second is more indicative of what characterizes the term "nocturne," and thus expresses its meaning more clearly. In this second example, also, although the definition is paraphrased, the source is given. With definitions this is a desirable practice, and for direct quotations citation of sources is a must. A research plan that is open-ended Delimitations will rarely result in the most expeditious accumulation and interpretation of data. In their understandable enthusiasm and zeal to make a significant contribution to the world's knowledge, research tyros

Research in Music Education

70

who have a study that is too broad frequently delimit their study even further than the subproblems require. At this point a distinction should be made between ''limitations " and ' 'delimitations. " Limitations represent a circumscription imposed upon a researcher due to external circumstances. For example, certain documents researchers need may be unavailable because they have been destroyed by some catastrophe, are lost, or are inaccessible because they are classified as confidential material. Delimitations are circumscriptions the researchers themselves place on their study . They are necessary to eliminate ambiguity and to exclude certain infen-ed items that the researchers do not intend to include. Delimitations are cognitive, conceptual, or what Statements, and often include a rationale for their inclusion. Some factors that might determine the extent of an investigator's delimitations include time and money available, personal qualifications to conduct the particular research, and availability ofspecial equipment or tools needed for the project. One of the most important of these is the element of time. This is not to imply that researchers should delimit their topic so much, ostensibly because of the pressure of time, that their design virtually is emasculated before the study has begun. The delimitations must be those that will enable them to organize their study in such a manner that they can effectively and efficiently obtain as complete data as possible. By contrast, there are not many occasions when researchers will find it necessary to expand their topic. Such a dilemma may suggest that the topic is too insignifiCant to be implemented. Delimitations may be negative-prescribing areas in which a study will not venture, thus giving investigators a better idea of what they do hope to effect-r positive-succinctly and clearly circumscribing limits of the study from an affirmative standpoint. Positive delimitations are not used as frequently as those stated negatively because it is easy to confuse them with objectives of the study.

Organizing the Research Proposal

71

Let us consider these delimitations in reference to the above hypothetical problem statement: (1) No attempt will be made to compare editions of the nocturnes of John Field and Frederic Chopin. (2) The moods pervading nocturnes are melancholy and dreamy . (3) The only system of analysis to be used in this study is the one developed by Heinrich Schenker. (4) This study will be delimited to a stylistic comparison of the eighteen nocturnes for piano by John Field and Frederic Chopin. No attempt will be made to compare the relative worth of the nocturnes of Field and those of Chopin. Of these delimitations, the first is viable because the researcher does not intend to compare the various editions to determine whether they correspond to what the Composer intended in the original version. The second is vague and would be better omitted because a positive delimitation is usually more difficult to conceptualize, and this one is a good example of the difficulty a researcher faces in providing a meaningful delimitation. The third delimitation is inaccurate because the Schenkerian approach has been modified by many individuals, including Felix Salzer and Adele Katz. Furthermore, it is more of an approach than a system for musical analysis. There would not be many occasions when a researcher would want to hew the line in strict fashion with only one approach: the complexity of many musical compositions require that an analysis use aspects of different approaches. The fourth statement, although verbose, expresses well the idea that the researcher is not trying to determine whether one composer's music is better than that of the other. Although this last statement may appear to be an oversimplification, and even a possible re-

Research in Music Education

72

dundancy, such lucid delineation is to the advantage of the investigator. Almost all research is Basic Assumptions predicated upon certain principles or propositions for which no proof is necessary or, in some instances, even possible. Researchers are not expected to prove their basic assumptions, but they should give a rationale for them. They merely present precepts that, according to Rummel, usually are "axiomatic in that they are propositions that virtually every reasonable Person is ready to adopt but which cannot be proven. "I2 It is not always necessary to incorporate basic assumptions in a research design. A trend toward omitting them has become evident to this writer. In many studies assumptions may be unnecessary, and to fabricate false assumptions just to meet a criterion is pointless. However, if they are used, they should be cognit?ve or conceptual Statements that are relevant to the study. Once again, refer to the hypothetical problem statement above as a decision is made regarding these basic assumptions: ( I ) It is assumed that the nocturnes of Frederic Chopin are more melancholy than those by John Field. (2) It is assumed that the first printed editions of the nocturnes by John Field and Frederic Chopin represent the true intention of the respective composers. (3) It is assumed that John Field was a less proficient performer on the piano than was Frederic Chopin. (4) It is assumed that a detailed analysis of the nocturnes of John Field and Frederic Chopin will enable a pianist to perform them more effectively . The first assumption is invalid and should not be included

Organizing the Research Proposal

73

because of a bias toward Chopin, suggesting that his troubled life caused him to write music that was more melancholy. In addition, the mood of the compositions being compared is not even under consideration in either the problem statement or the subproblems. Mood could be a topic in itself. The second is a very important assumption because there will be many occasions when one is unable to obtain an original manuscript to compare that version with the first printed edition. Furthermore, in the proposed study there is no attempt to determine whether the first printed editions of each composer are to be compared with the original manuscripts to determine what changes the composers may have made from the time the manuscripts were completed and the first editions were printed. The researcher proposes to deal only with the first printed editions of each composer. The third assumption also is invalid and biased because Field is not as well known as Chopin. Furthermore, performing ability of the composers is not a focus of the study or even relevant to it. Such a study would be difficult to objectify at best. The last assumption is valid and really serves as the focus of the study. By examining in detail the music of both composers one should be able to understand the works better and thus perform them more effectively. A hypothesis is a shrewd guess of the outcome of a study. Like Basic Hypotheses basic assumptions, basic hypotheses are not necessary to include in all research designs, but they appear more frequently than do assumptions. They rarely appear in historical or philosophical designs. If they are present it would be after the research is completed and the researcher has formulated hypotheses based on its results. Some aesthetic designs call for them; others do not. It is very rare to find descriptive and experimental designs that do not include them, as already noted. Basic, sometimes known as "working," hypotheses are indications of a researcher's guess as to the consequences of a study. Setting forth the basic theory or rationale underlying the

Research in Music Education

74

hypotheses prevents the appearance of false or wild hypotheses. Hypotheses remain tentative until the research is complete, when they are either confirmed or rejected. There are various types of hypotheses, but two are useful for most music-education research: the research, also known as deductive, declarative, or positive, and the statistical, null, or negative. The research hypothesis is stated in form of a difference that is expected. A researcher might use this formula: A > B, to test whether A is indeed greater than B. If it is, then the hypothesis is confirmed or sustained. On the other hand, if A does not prove to be greater than B (A < B), then the hypothesis is rejected. The statistical hypothesis simply states that no differences may be expected between two or more variables when measured statistically . Although this procedure usually results in a direct testing of the hypothesis, it nonetheless is constructed with the expectation that it will be rejected. Rejection then re'sults in affirmation of the research hypothesis. Using the formula A # B, the researcher hopes that there will be a difference so the null hypothesis can be sustained. A null hypothesis is not accepted; rather, one fails to accept it. It may seem pointless to state something negative when it might appear easier to do so positively, but statistically a null hypothesis is easier to handle because it is predicated on the Gaussian or bell-shaped curve. (Further explanation can be found in Chapter 6.) To reiterate, a hypothesis should be stated clearly and concisely. When so formulated, it is easier to test, thus fulfilling the purpose of the research more readily. To put it another way, a basic hypothesis is an unbiased statement of anticipated conclusions. Note that hypothesis is not synonymous with conclusion. After data have been accumulated, a conclusion might, of Course, be coincidental with a hypothesis formulated at the initiation of the research design. Obviously, any bias will tend to prevent not only the accumulation of valid data but also the deduction of sound conclusions.

Organizing the Research Proposal

75

Let us examine once more the abovementioned hypothetical statement to determine which of the following statements represent viable hypotheses: (1) It is hypothesized that, since the term "nocturne" is used by both John Field and Frederic Chopin, no technical difference between their nocturnes will be evident. (2) It is hypothesized that, because of his national background, the nocturnes of Frederic Chopin will be more somber than those of John Field. (3) It is hypothesized that melancholy compositions will be more highly developed stmcturally than will be pleasurable compositions. (4) It is hypothesized that there will be no difference in the results obtained irrespective of the system of stylistic analysis used. The first hypothesis is not valid, or testable, because each composer writing under the rubric of "nocturne" obviously is going to express himself in his own way. Furthermore, there is no technical limitation implicit or explicit in the term "nocturne. " The statement "no difference" obviously designates this as a null hypothesis. The second hypothesis is invalid and untestable, because national background and mood are not common elements. To state that the music of composers of any one country is more or less melancholy is irrational because it has no basis. Since the word "more" is used in this hypothesis some specific direction is indicated, and the hypothesis would be labeled as "directional." If the statement simply stated that there was a "difference" it would be nondirectional, and the researcher would then have to determine through a comparison of the variables whether the judgment should be "more " or "less. " Research and null hypotheses should not be mixed up in any one

Research in Music Education

76

design. In other words, one hypothesis should not be positive and the next one negative, or vice versa. The third hypothesis places emphasis on mood and structure. There does not necessarily need to be a relationship between them, so this hypothesis also would be very difficult to test. In the final hypothesis the focus is on the system of analysis used, which obviously is a fallacy. One of the important decisions a researcher has to make is whether any one system or approach of stylistic analysis will bring more valid results than any other. The upshot of this discussion is that none of the hypotheses above is valid and testable; all are false. Not many studies of the type suggested above require hypotheses. However, hypotheses need to be formulated and tested in experimental research, as will be Seen in Chapter 6. Up to this point in the reNeed or Significance search design the sections do of the Study not contain much narrative. But in the need section investigators must clearly Set forth why they feel their proposed study is needed and how the results might possibly be used. It is their task to convince the reader that the study will result in a signifiCant contribution to the literature of the field. Sometimes this section of a research design is called "Significance of the Study," but regardless of the designation, investigators must avoid giving, on the one hand, the impression that here at last is research that will resolve the most vexing problems in music education and, on the other, an aura of apology. For instance, the use of such equivocal phraseology as "interest of the writer, " "encouragement from the writer's Peers," and "the writer's desire for musical growth" is not very persuasive. Such euphemisms are redundant, to say the least, and are an indication of intellectual immaturity. To lend validity to their assertions researchers should use citations from the literature or from individuals who share their convictions about the need for such a study. The more convincing the evidence for need is, the less likely a reviewing panel will say "So what?" after reading the research design. Indicate the signifi-

Organizing the Research Proposal

77

cance of the proposed study for music education, for the music teacher in the field. This evidence may be based on critical thinking or obtained from the literature. The presence or absence of corroborating and supporting declarations, moreover, may or may not be sufficient affirmation that the research is needed. To be significant a proposal must rest on its own merits, with or without the substantiation of others. Sometimes researchers will include a forthright and convincing account of what caused them to select the proposed topic. Sometimes referred to as "Incidence of the Problem, " this information, if needed, should be included in the need or significance section. One of the most important Related Literature steps in the preparation of a research design is a review of related literature. There are several reasons: the review can help the researcher (1) avoid duplication of efforts; (2) delimit the research problem; (3) determine what areas need further investigation; and (4) discover new approaches, methods, or insights into a problem. In the related-literature section investigators include a brief historical resume of their particular area of research, listing some of the most important sources of pertinent information. They should, in addition, include concise accounts of other concepts that may be relevant to their topic, such as philosophical, psychological, or sociological implications. In a recent publication this writer stressed the importance of a thorough review of the literature because researchers undoubtedly will be the experts in their specific area of research by the time they complete the study. Without a careful examination of the literature investigators will not know what has been researched, and will thus find it difficult to know what remains to be done.I3 The researcher presents a bnef Summary of the significant

Research in Music Education

78

points found in each source reviewed along with the relevance of each one to the study. While it is important to present enough information on each source to give an adequate picture of its relevance, verbosity must be avoided lest this section balloon out of proportion to the other sections of the research design. The information needs to be presented from both positive and negative points of view. Researchers should not hesitate to point out materials, for instance, that have little or no pertinence to the study. On the other hand, they may have omitted some items purposely, and the uninformed reader needs to know why. As already stated, researchers are expected to survey the literature of their field completely and carefully. This significant step will prove beneficial when they organize their bibliography for the final report. Citations, when fitting, are included in this section, with appropriate comesponding footnotes, since they help verify that the investigator is thoroughly familiar withthe literature. The onus is upon the researcher to determine which materials are pertinent and which are not. Two parts normally comprise the Methodology methodology section: "Sources of Data" and "Treatment of Data. " In the first part the researcher indicates where and how the information wili be obtained. This should be provided for each subproblem, and represents the how phase of what is proposed in each subproblem. The researcher includes the following in the sources-of-information section: (1) literature or other sources to be used; (2) kinds of data needed; (3) amounts of data needed; (4) when data should be sought, if a time factor is significant; ( 5 ) instruments with their validity and reliability coefficients, and equipment or materials to be used for data-collection; and

Organizing the Research Proposal

79

(6) personnel assisting with the accumulation of data. The focus and scope of the research problem will determine the number of these components, and if others are needed. Cook and LaFleur point out the importance of describing research procedures clearly and concisely, regardless of the techniques used.14 In the sources section a researcher will frequently pose certain questions for which answers are being sought. These need to be clearly delineated. The answers will not be complete until all the data have been received. Questions to be posed on a questionnaire or in an interview or statistics of experimental designs also appear in the sources section. When a jury is being used, researchers must indicate why these particular jurors were selected rather than some others, specifying the criteria for their selection. Once the procedure for collecting data has been established, the researcher has not only a clear and concise guide for conducting the study but also a plan that may be followed by another investigator who might desire to replicate the research. In the treatment-of-data section the researcher indicates what will be done with the information obtained in each subproblem. The data, biographical and historical for instance, may come from literary sources, in which case the researcher states that the information will be presented in "chronological, narrative form." A research study based on an analysis of music compositions should include a sample of materials that have been analyzed. The analysis could be either in the treatment section or in an appendix. All references to appendix material should appear in the appropriate place in the context. Procedures for handling questionnaire data, such as key punching, should be listed in this section also. When historical material is presented researchers must indicate how they will determine the authenticity and credibility of documents or statements they wish to include. The sole purpose of the methodology section is to let a reader know where data will come from and what will be done with

Research in Music Education

80

these data after they have been obtained. Methodology is the most significant section of a research proposal, and as such requires very careful critical thinking and serious preliminary investigation of the problem. Known also as "Preliminary " or Bibliography "Tentative " bibliography, this section contains references that the investigator considers to be most important. No attempt should be made in the research design to incorporate all reference material~.However, a requirement for the Alternative Ed.D., to be discussed shortly, is that an amotated account of bibliographical materials be given, but this is normally not done in the Ph.D. and traditional Ed.D. As the study Progresses additional sources undoubtedly will be located. The quest for data is an ongoing process, but the investigator needs to establish limits. In the final report is included not only ,the most current information but also the most recent bibliographical items relevant to the study. This bibliographical process continues right up to the time of the completion of the final draft of the dissertation or thesis. It may prove beneficial in some designs to separate bibliographical items into primary and secondary sources. The former, of Course, refers to original or firsthand accounts of information, such as personal letters or data reported as the result of an experiment. A secondary source contains items that are not original with the one reporting them. For example, most books on the history of music would be considered secondary sources. Although the authors presumably used primary sources of data, their reporting of them results in accounts that are secondary. (This writer has difficulty on many occasions in making a clear distinction between primary and secondary sources, and therefore prefers that the bibliographical items be listed by category, for example, books, periodicals, unpublished materials, etc.) In this section of the proposal sources are presented in standard

Organizing the Research Proposal

81

bibliographical form. Researchers must be careful to include the most recent edition of a publication; materials must be as up-todate as possible. This does not mean that older sources are less valuable, but rather that investigators are obliged to keep informed of both past and present in their area of concentration. There are instances when older sources still are the definitive ones, so they should be used when needed. This section gives the inPersonal Qualifications vestigator 's education, degrees, Courses taken, and professional and other experiences. In addition, writing and research may be listed in some research designs, but these are optional. The import of these items is evident; their main purpose is to show that the researcher has adequate preparation and background to pursue the proposed research successfully. The appendix normally contains inforAppendix mation to be used in implementing or clarifying the proposed research. Reference to these items, as already observed, is made in the appropriate place in the context of the design. Preliminary questionnaires and proposed charts for musical analysis are typical materials to be included.

Format for Nontraditional Doctorates Since the Alternative Ed.D., the D.A., and other "nontraditional" doctorates are largely tailored to the individual needs of students, the research proposal for these designs is less rigid than that for the Ph. D. and the traditional Ed.D. However, many components required for the Ph.D. or Ed.D. also appear in these proposals . The Alternative Ed.D. proposal contains the following: problem Statement; subproblems; significance of the study, including

Research in Music Education

82

relevance for education; methodology; annotated bibliography; and relevant personal and professional experiences. The proposal is not to exceed ten pages, exclusive of annotated bibliography and personal and professional experiences. The most obvious departure from the Ph.D. or traditional Ed.D. is that of annotated bibliography. In the Alternative Ed.D. the annotated bibliography replaces the related-literature and bibliography sections found in other proposals. (Although no Alternative Ed.D. doctorates had been completed in music education at New York University at the time of this writing, Ruth Coron, in accord with the interdisciplinary nature of the degree, completed "Interdisciplinary Teaching in the Arts: A Videotape Presentation, " in 1974.15 Coron, a visual-arts teacher, was joined by team members who were specialists in theater arts, movement, and music to illustrate the interrelationships among lines, textures, colors, shapes, rhythms, and spaces using demonstrations that were recorded on vide'otape. A written document accompanied the videotape.) The Doctor of Arts (D.A.), also is flexible and tailored to the individual needs of students. Its proposal may contain all or any combination of the components found in the Alternative Ed.D. or Ph.D. and traditional Ed.D. research proposals. The end product of the D.A. is a research essay, tobe developed in accord with the Student's practicum, internship, or apprenticeship. Proposals for the D.M.A. and Mus.D. normally follow procedures and format similar to those for the Ph.D., incorporating modifications necessary to accommodate musical skills and analytical techniques.

Deficiencies in Research Designs The foregoing pages have stressed factors to be incorporated in a sound research design. Yet, the fact remains that knolvledge of

Organizing the Research Proposal

83

what should be and what actually is contained in a research design do not always coincide. Since this chapter has dealt largely with the research-design plan in effect at New York University's School of Education, Health, Nursing, and Arts Professions, a reasonable postlude is an account of deficiencies observed in designs organized according to this format. This is not to imply that proposals at New York University are any better or any worse than those of comparable institutions. This writer's evaluation of research designs completed elsewhere shows a general consistency indicating that common areas of deficiency in any one institution will very likely also be found at another. In one six-month period eighty-eight research designs were examined by a committee appointed by the dean of New York University's School of Education, Health, Nursing, and Arts Professions (See Figure 3). Subsequent evaluations have revealed a similar Pattern of the weaknesses listed here.I6 Methodology, not unexpectedly, proved to be an overwhelmingly consjstent area of deficiency. The term is analogous to data sources and treatment procedures, which were discussed earlier in this chapter. Weaknesses under this category include: failure to list methodology for all subproblems; failure to give procedures for testing hypotheses; use of general terms, such as "analyze, " "survey, " and "research, " without indicating their specificity; lack of evidence regarding reliability of research instruments and techniques; no provision included for objectively evaluating data; missing selection criteria for sources; and failure to list procedures for answering questions posed. The most consequential inadequacies under the category of "Assumptions" resulted from investigators' failure to defend or justify their assumptions. In other instances, items listed here should have been included elsewhere. For example, when investigators "assume' ' their project is significant, they should substantiate their assumption under "Significance of the Study," not under "Basic Assumptions. "

Research in Music Education 84 Figure 3 Deficiencies Reported in Research Designs

Area of Weakness

(%) Containing Weakness

Methodology Assumptions Hypotheses Suitability for Degree Delimitations Subproblems Definitions Related Literature Significance of Study Format Problem Bibliography

Likewise, factors often listed as hypotheses should actually have been presented as assumptions or in some other appropriate area of the design. In addition, hypotheses were given that bore no relationship to the subproblem or that were not subsequently developed in the design. A major weakness appears to be the enumeration of hypotheses that are not testable. When "Suitability for the Degree" was indicated as a deficiency, the implication was that an Ed.D. design was offered for a Ph.D. proposal, or vice versa. As noted in Chapter 1, the difference between the two usually is that the Ph.D. is regarded as pure research, with readers expected to make their own applications of the research results, and the Ed.D. as action research, with the investigator incorporating uses of the data into the study. Often the general topic chosen rnay be suitable for either degree,

Organizing the Research Proposal

85

but the formulation and implementation of the design rnay well determine whether the study will be for the Ed.D. or the Ph.D., when such options are available. The investigator who presents delimitations convenient to manipulate rnay be guilty of altering the direction the research normally would take because vital factors rnay have been excluded. Another common weakness is failing to give reasons for specific delimitations. Invalidated Statements naturally are suspect. Failure to include all the subproblems necessary to resolve the general problem of an investigation is, of Course, dereliction in the organization of a design. This oversight rnay be due either to carelessness or to intellectual ineptitude. Another deficiency is that of confusing actual procedures with the formulation of subproblems. Subproblems actually are indications of areas to be investigated and as such require no designation at this point of how the research will proceed. Definitions that lack delineation and specificity appear to be a common weakness in designs. Occasionally researchers will fail, also, to state how their definitions vary from generally accepted usage. Inadequacies in the section entitled "Related Literature" essentially consist of failure to indicate how the materials relate to the study. Perhaps the researcher should give more consideration to the possibility that a superficial or incomplete review of the literature rnay cause a proposal to be disapproved by a reviewing panel. Reference has already been made to this, but such inadequacy is inexcusable if for no more compelling reason than the ethical and moral principles of scholarship involved. One of the most recurrent deficiencies in the significance-ofthe-study section is the inability of a researcher to explain why the proposed study will be important. The assumption that the absence of research in a specific area is enough to justify it is fallacious. Included under the area of "Format" are such miscellaneous

Research in Music Education 86 items as misspellings, lack of pagination, disjunct presentation, and indifference to accepted procedures for conducting scholarly work. Although they may appear to be of minor import, any study that treats these factors in a cavalier manner can hardly be classified as good research. This writer is appalled at times by the apparent lack of concern on the part of some doctoral students who disregard correct format for footnotes, spelling, and conjunct presentation. Graduate students should not expect their advisers to remind them of all these details. If they are incapable of the particularity required for adequate proofreading, they should seek professional help for this chore. Although failure to state a problem precisely is seemingly an imocuous weakness, it is difficult to comprehend how any problem can be successfully pursued without a clear and concise formulation of the research. Another deficiency is presenting the problem as a prejudicial instead of an unbiased,*proposition. The tendency to present a topic in hortatory manner may be anticipated when an individual is exuberant about the research but uninitiated in its techniques. Finally, weaknesses in the bibliography category stem largely from researchers' failure to present references in a form generally accepted for scholarly work and from the inclusion of many items relatively insignificant to the topic. The danger is that research tyros may believe the inclusion of a large number of references will suffice to indicate that they have acquainted themselves with the literature of their field. The result may be that they will neglect to include a few works that are essential to the study at the expense of numerous items of minor importance. Whatever it may be called a research design or dissertation proposal serves a very important function. It is the researcher's plan or blueprint, as it were. A well-organized and carefully prepared proposal can make the implementation of a research plan easier and the interpretation of data much more meaningful. A most important step in research plaming is a review of litera-

Organizing the Research Proposal

87

ture in the field. Suggestions for obtaining bibliographic information follow in the next chapter.

Problems for Review und Discussion 1. What is the function of a research design? What changes are permissible in a design as the actual research proceeds?

2. Compare the research-design format recommended for use at your institution with the one discussed in this chapter. Note particularly elements that are common to both. 3. Why is the "Definition of Terms" section of a research design so important? 4. Differentiate between positive and negative delimitations and give an example of each.

5 . What is a basic assumption? How does it differ from a basic hypothesis?

6. Differentiate between a deductive and a negative hypothesis and give an example of each.

7. Why is a review of related literature one of the most important steps in the preparation of a research design?

8. Compare the deficiencies in research designs listed in this chapter with those apparent in projects presented by members of your class. 9. Select a tentative research area and pursue it to formulate a title, problem Statement, subproblems, definitions of terms,

Research in Music Education

88

delimitations, basic assumptions, and basic hypotheses i n the manner discussed in this chapter.

Supplementary Readings BORG,WALTERR., and MEREDITH D. GALL.Educational Research: An Introduction, 3rd ed.; New York: Longman, 1979, Chapter 2. BURKE,ARVIDJ . , and MARYA. BURKE.Documentation in Education, 4th ed., rev. New York: Teachers College Press, 1967, Chapters 1 , 4, 14. and 15. GARRETT,ALLEN M. An Introduction to Research in Music. Washington, D.C.: The Catholic University of America Press, 1958, Chapter 9. Understanding ReHARDYCK, CURTIS,and LEWISF. PETRINOVICH. search in the Social Sciences. Philadelphia: W. B. Saunders, 1975, Chapter 2. HILLWAY, TYRUS.Introduction to Research, 2nd ed..Boston: Houghton Mifflin, 1964, Chapters 9 and 10. HOPKINS,CHARLES D. Educational Research: A Structure for Inquiry. Columbus, Ohio, Charles E. Merrill, 1976, Chapter 5. KERLINGER, FREDN. Foundations of Behavioral Research, 2nd ed. New York: Holt, Rinehart and Winston, 1973, Chapter 2. L A B O V I TSANFORD, ~, and ROBERTHAGEDORN. Introduction to Social Research, 2nd ed. New York: McGraw-Hill, 1976, Chapter 2. MASON,EMANUEL J . , and WILLIAM J. BRAMBLE. Understanding und Conducting Research: Applications in Education und the Behavioral Sciences. New York: McGraw-Hill, 1978, Chapter 3. MCGRATH,G. D., JAMESE. JELINEK,and RAYMOND E. WOCHNER. Educational Research Methods. New York: Ronald Press, 1963, Chapters 5 and 7. PHELPS,ROGERP. "The Doctoral Dissertation: Boon or Bane?" College Music Symposium, Fall 1978, 18,2:82-93. RUMMEL, J . FRANCIS. An Introduction to Research Procedures in Education, 2nd ed. New York: Harper and Row, 1964, Chapter 3. SAX,GILBERT. Empirical Foundations of Educational Research. Englewood Cliffs, N.J.: Prentice-Hall, 1968, Chapter 5.

Organizing the Research Proposal

89

TRAVERS, ROBERTM. W. An Introduction to Educational Research, 4th ed. New York: Macmillan, 1978, Chapter 4. TucKMAN, BRUCEW. Conducting Educational Research, 2nd ed. New York: Harcourt Brace Jovanovich, 1978, Chapter 2. VAN DALEN,DEOBOLD B. Understanding Educational Research: An Introduction. New York: McGraw-Hill, 1979, Chapters 7 and 8. WHITNEY, FREDERICK L. The Elements of Research, 3rd ed. Englewood Cliffs, N.J.: Prentice-Hall 1950, Chapters 5 and 6 . WISE, JOHN E., ROBERTB. NORDBERG,and DONALDJ. REITZ. Methods of Research in Education. Boston: D. C. Heath, 1967, Chapter 2.

Obtaining Resource Materials

91

Efficient Library Use

4 . Obtaining Resource Materials Any research depends to a certain degree on the utilization of bibliographic techniques. Some types of investigation require more extensive source materials than others. For example, historical, philosophical or aesthetic inquiry or research will more likely rely heavily on data to be found in libraries and archives than will experimental or descriptive research. It is in the library that researchers normally receive preliminary information that corroborates the feasibility of pursuing the proposed topic. It is here also that they procure much of the information they eventually use in their preliminary and final bibliographies. Yet it is paradoxical that many graduate students, embarking on the initial Stages of thesis or dissertation planning, are still unfamiliar with some of the most fundamental techniques of efficient library use.

It hardly seems necessary to suggest that researchers should first acquaint themselves with the general floor plan and holdings of the library. This is not to suggest that they will need to examine every listing in the card catalog before selecting a topic. Instead, they should become familiar with the library's card catalog, general and special collections, reference rooms, and facilities for the reproduction of materials. General collections include books, periodicals, government documents, and other publications, both of recent and of older vintage. Special collections might consist of such items as writings, diaries, programs, letters, manuscripts, and other personal effects of an individual or group. For example, a researcher investigating the role of music in the life of Georgia poet-musician Sidney Lanier (1842-188 1) might examine holdings of the Lanier Room at Johns Hopkins University in Baltimore, Maryland. Uncataloged materials in special collections are often a lucrative source of information. For example, one of this writer's doctoral students found uncataloged materials in the library of the Lincoln Center for the Performing Arts in New York to be a fruitful source of information on Olga Samaroff Stokowski.' The Stokowski materials include letters, programs, concert reviews, citations, curriculum guides, and miscellaneous memorabilia. The reference room contains standard sources of general information, such as encyclopedias and dictionaries of all kinds. Finally, because an individual may wish to have an item reproduced by some photographic process rather than copy it by hand, especially if the material is extensive, it would be well to know what technical processes are available at the library, how much they cost, and how long it will take for duplication. A general overview is necessary for investigators to determine how many of the source materials are housed in a given library, and they will also need to learn whether they have to look else-

Research in Music Education

92

Obtaining Resource Materials

where for appropriate items. Should this be necessary, they still must find out where such materials may be found. Professional colleagues, library staff members, and advisers are the most likely sources of such information. They eventually may direct the researchers to another library to repeat the Same procedures of acquainting thernselves with its holdings and facilities, examining the materials, and so on, until the needed information is found. Researchers should check Subject Headings2 before consulting the card catalog in a library that uses the Library of Congress System. A comparable qompilation is available for libraries using These lists are helpful to the Dewey Decimal Clas~ification.~ researchers because they specify what subject headings are used in the card catalog. Another good source of information is Books in Print .

Library Card Catalog The card catalog undoubtedly is the most important single source of preliminary information for researchers. Cards may be prepared by the library staff or they may be published by the Library of Congress or commercial book distributors. Each card contains a wealth of information that sometimes is overlooked by the research tyro. Figure 4 is a reproduction of a catalog card for a recent publication in music. It is referred to as an author or main entry card. The identifying numbers correspond to the following: (1) library call number, which identifies where the book is shelved; (2) author's complete name; (3) date of author's birth (and death in the case of an author who was deceased at the time the card was prepared, or whose demise subsequently has been recorded by a cataloger); (4) title of publication;

93

Figure 4 Library Catalog Card

Bureau of P blications. T e a c h e r s College, Columbia University, 1928. ,New ~ o i kAYS , P r w , 1$2, X, 107 p.

I

22 Cm.

Reprint uf the 1948 eä., lssueä In serles: Teachers College, Columbia Uuiversity. Contriibutlons to diicatlon, no. 885. + Orlginally presented a s the author's thesis, Coliirnbla. Bibllography : p. 101403. 1. Choral music-ünlted States. I. Tltle. 11. S e r i e : Columbla Unlversity. Teachers College. Coutrlbutious to educatlw, UO. 885.

13 20 16

ML1511.N5C5 1972

784'.1

I S E N @-404-55@&2 Libraiy of Congress

73 Pi

t

J

21

14 75-176698r-15 M-4Rc-18 MN

t

( 5 ) edition number often shown here (in this example the edition number listed in (21) is that of the publisher); (6) place of publication; ( 7 ) publisher; (8) date of publication; (9) nurnber of pages in preface and in main body of book; (10) height of book; (1 1) page references to bibliography; (12) rlotation of "tracings" or subject or other headings under which the book is listed in the card catalog; (13) Library of Congress cnll nurnber;

Research in Music Education

94

( 14) Dewey classflcation number;

(15) Library of Congress card nrimber; (16) publisher of card; (17) code for card; (1 8) Library of Congress computer-system designation; ( 19) Libraq of Congress special-area designation; (20) International Standard Book Number; and (21) the series to which the book belongs normally goes here. (This example is an exception. The reprint is not part of the senes as such. The original volume would have shown series in parentheses at end of collection .) The library's call number (I) is inserted by the libraq staff. In addition to the main-entry card shown in Figure 4 the card catalog contains additional entries for each publication. Cards containing identical inforrnation are filed in appropriate places of the card catalog according to the additional information typed at the top. The title card lists the title of the publication, and the subject card identifies the subject, as shown at the bottom. Subject headings may appear typed in red or in black capital letters (sometimes both), depending on local preference, to distinguish them from other added designations. In addition, all libraries maintain a file known as a shelflist, containing, in call-number order, one card for every item in the library, and usually the number showing "holdings, " that is, volumes, number of copies, etc. (In many libraries, the main entries also show holdings.) The shelflist, normally used only by catalogers and others in technical Services, is often available for use under the supemision of library personnel. Ir is important for the researcher to copy everything in a call number e.xactly, line for line as it appears on the card.

Obtaining Resource Materials

95

Because the book listed in Figure 4 is a repnnt, two dates of publication are cited, the original and the reprint dates (8). Although the Same city of publication (6) appears for both versions, it is indicated in two places, as are the two different publishers (7). Of some interest to researchers is (9), number of pages in preface and in main body of the book. Bibliography references (1 1) could lead researchers to additional sources of materials, and (12) tracings may be helpful in locating items that might be overlooked under different categories. Certain items have little or no significance for a researcher and are primarily of interest only to members of the library staff. These are height of book (10), Library of Congess call number (13), Dewey classification number (14), Library of Congress card number (1 5), publisher of the card (16), code of date of card (17), Computer designation (1 8), special-area designation (1 9), and International Standard Book Number (20). To a librarian, height of a publication can be significant because items in music and art are bound in a variety of sizes and shapes. To pose a very practical question: should the height of the adjustable shelves found in most libraries be altered to accommodate numbers of large musical anthologies, Scores, and art portfolios, thus wasting precious space? Or should "oversize " resources be housed separately from smaller books? The staff members of each library must answer these questions in accordance with the individual situation. Researchers, however, should be aware that shelving may be divided by size. Cards for large-sized materials bear some type of Such notation within the call number, frequently the Symbol an indicator must be included on the call card or slip submitted by a researcher. In addition, items may be shelved in a section of the library where they would not normally be expected to appearfor example, in a special collection, or series, or with a related area of scholarship. In decentralized libraries researchers should first examine the union or "public" catalog. The location of the material is usually

+.

Research in Music Education 96 noted at the top of at least the main-entry card; for example: Music Library, Education Library, Science Library. A book on music education, for instance, might be in the education rather than the music collection. Investigators who examine only the music card catalog might not find the desired publication listed there and thus assume that it was unobtainable in that library. Returning again to information included on a catalog card, note that the author's complete name, the correct title of the publication, the edition, the publisher, and the place and date of publication are all significant information to be included in a bibliography. The author's dates become significant when one is concemed with writings from a certain historical period, as would be true for many historical and philosophical research studies. Other headings, as shown in the tracings, can be very helpful in locating additional sources relevant to a topic. For example: an investigator knows that the name Davies was associGed with the psychology of music. By checking the subject heading "MusicPsychology" the researcher can find other materials on the Same subject. Although the classification of materials in American libraries may vary in detail from one library to another, most items on the Same subject will be classed by either the Dewey Decimal classification or the Library of Congress system. Many smaller college and municipal libraries use the Dewey system; larger college, university, and city libraries find the Library of Congress schedule to be more useful because the combination of letters and numbers and the larger number of classes makes it more flexible than Dewey. First published in 1876 by Melvil Dewey (1851-1931), renowned progenitor of sound U.S. library techniques, the Dewey Decimal Classification provides for the classing of published material~under ten major headings: 000 General Works 100 Philosophy and Psychology 200 Religion

Obtaining Resource Materials 300 400 500 600 700 800 900

97

Social Sciences Philology Natural Sciences Applied Sciences and Technology Fine Arts and Recreation Literature History , Geography , Biography .

Music entries are assigned numbers from 780 through 789, with these representative examples: Research General Principles ("Theory of Music ") and techniques Dramatic music and production of musical drama Sacred Music Sacred Music-Christianity Voice and vocal music Appreciation (analytical and program notes) Instrumental ensembles and their music Orchestra Keyboard instruments and their music Historical and geographical treatment String instruments and their music Violin Wind instruments and their music Oboe and English hom Percussion, mec hanical, electrical instruments Membranophones Catalogs, lists, critical appraisal of recordings .5

Research in Music Education

98

The Dewey classing scheme has been replaced in many libraries by the Library of Congress system, devised by Herbert Putnam (1861-1955) and published in 1902. This is especially true in libraries where a greater diversification is desirable because of large holdings. The system of classification devised by Putnam for the Library of Congress, one of the world's largest repositories of materials, is: A General WorksM Music Polygraphy N Fine Arts B Philosophy-Religion 0 Vacant, left for expansion C History-Auxiliary P Language and Literature Sciences Q Science D History and Topography R Medicine (except America) S Agriculture-Plant and E-F Arnerica Anima1 Industry G GeographyT Technology Anthropology U Military Science H Sociology V Naval Science 1 Vacant, left for expansion W, X, Y Vacant; left for J Political Science expansion K Law Z Bibliography and L Education Library Science. Oscar G. T. Sonneck (1873-1928) did the original scheme for music. Music entries are classified under three broad subheadings of the M classification system of the Library of Congress, as follows: M Music [i. e., scores] ML Literature of Music MT Musical Instruction and S t ~ d y . ~ Some large subdivisions within each of these areas are: M Music 5-1459 Instrumental Music 1495-2199 Vocal Music

Obtaining Resource Materials

99

ML Literature of Music 159-3790 History and criticism 1100- 1354 Charnber and orchestral music, band (military music) 3800-3920 Philosophy, physics, psychology, esthetics, etc., of rnusic MT Musical Instruction and Study 40-67 Cornposition; rhythm, melody, harmony, counterpoint, etc . 170-810 Instrumental technics; for Organ, piano, string, wind, and plectral instruments 820-949 Singing and voice culture 855-883 Special technics; chanting, sight singing , choral singing , etc .' Each of these is further divided into small topics. From the extensive number-identification system noted above it is easy to See why there might be occasional discrepancies in the classification of publications. The classification Systems, both of Dewey and Putnam, are suggestive only. It is the prerogative of the individual cataloger to assign a specific number to each item according to the library's present holdings. Note that the general classifications should fall within these basic patterns of headings. It would be unlikely to find materials relating to music instruction (MT classification) listed under ML (Music Literature). Some music libraries, moreover, use a scheme of classification for musical scores based on a decimal system devised by George S. Dickinson (1888-1964) though often modified by individual libraries to accommodate local conditions. The Dickinson-Columbia schedule includes these classifications for musical scores: 00-09 Collections, general and instrumental 10-69 Instrumental music (by medium) 70-99 Vocal m u s i ~ . ~

Research in Music Education

100

Since the trend in library science today is toward Open Stacks or shelves, students familiar with classification systems can go to the appropriate area of a library and examine materials firsthand. Another advantage of becoming familiar with these schemes is that investigators rnay locate items appropriate to their musical topic that are classed under another area heading. The appropriate classification of materials in a library obviously is not the responsibility of music researchers, but an understanding of the principles of classification can be useful to them in locating materials relevant to their research. While most of the remarks above relate to published materials such as books, researchers rnay have occasion to examine musical scores, which, unfortunately, have less uniformity in their systems of classification. Often scores are merely placed in portfolios and arranged alphabetically by title and by composer, conductor, or performer. In other instances they rnay be found arranged according to some modified classification system. Some researchers rnay have occasion to use disc and tape recordings that are housed in special collections of a library. Library of Congress cards are available for many of the disc recordings, and this, of Course, makes the search much easier. The LC card numbers rnay be found in the discography section of each issue of Notes, the quarterly journal of the Music Library Association .9 Some libraries, especially those in which records circulate, maintain a separate card file for disc and tape recordings because of the ease with which users rnay find the titles, which are conveniently listed by composer. Another reason is that records, as Shank and Englebrecht correctly state, wear out easily through handling and often will not be replaced, thus making it logical to maintain a separate catalog for them.1° Some libraries now do not permit disc recordings to circulate, instead keeping them as "master copies" and making audiotapes or cassettes that circulate or are reserved for use in the library's listening room.

Obtaining Resource Materials

101

It should be clear to the reader by now that the card catalog of a library can be a most valuable source of preliminary information. Regrettably, many researchers do not use it as effectively and efficiently as it should be.

Suggestions for Note Taking Even before detemining what materials they want to examine, researchers should ascertain which system of note taking will best suit their needs. Actually, a uniformity in procedure is more important than the plan followed. The most widely used method involves making notations on blank or lined file cards, usually 3" X 5" or 4 X 6" because of the ease with which they rnay be handled and filed, preferably in an easily carried file folder or box. Other researchers prefer to place inforrnation on blank pieces of paper 8%" X 11" or smaller, and then file them in a large manila envelope." The advantage of this system is the ease in keeping material from a single source on one or two pages instead of on several filing cards, thus permitting facility of reference. (By the way, manila envelopes containing a string loop fastener usually are more durable than those with flexible meta1 clasps.) A third method is to make notations in a looseleaf or spiral notebook. Although the looseleaf notebook makes it easy to arrange materials, it frequently proves to be less durable because of inconvenience in handling. Spiral notebooks have the disadvantage that pages rnay be easily detached; it would be well to carry a few paper clips to attach loose sheets to the notebook so they will not be misplaced before they have been inserted in an envelope or file box. Some researchers object to using spiral notebooks on the premise that pages detached have rough serrated edges that are difficuit to clip together, and there are aesthetic objections also to the appearance of these irregular sheets.

Research in Music Education

102

Whatever system they use, researchers must be methodical and consistent in their procedure and approach so they can obtain most of the information they need the first time they examine the materials. This is accomplished only with patience and practice. Smith is not alone in offering the advice that systematic note taking is accomplished only as a result of careful preparation, organized thinking, and resolute application.I2 Baughman suggests that color coding, either through using different-colored file cards or coloring the tops of file cards with crayons, may help organize material obtained from various sources.13 The data to be recorded on each card or sheet must be complete enough not only to be useful but also to save researchers from having to reexamine the same item for information overlooked the first time. They should record first the library call number for the item in the upper-left-hand Corner (also location, if more than one library is used), and other pertinent identifying information in case there is occasion to refer to the material'again. Next, every notation should list complete bibliographic information (author's full name, complete title, place, publisher and date of publication, and pages used). (How many times does a research tyro find it necessary to look up forgotten Page numbers for quotations?) Notes should be placed on only one side of a card or sheet and may consist of: (1) direct quotations, (2) paraphrase of material and (3) outline of important points. In order to differentiate them from other types of commentary, direct quotations are enclosed between quotation marks. Intellectual honesty demands that all items be quoted exactly as printed; this includes marks of punctuation as well. Mistakes in the original must not be corrected, but rather indicated with the Latin word [sic]in brackets, underscored because it is a foreign word. If extensive portions are to be quoted it would be advisable, in the interest of time and accuracy, to have them reproduced at a nominal cost by one of the photocopying processes now available in most libraries.

Obtaining Resource Materials

103

Materials should be paraphrased whenever possible to develop the facility for expressing in one's own words what one reads. This will rnake it easier to resist the temptation to present a report that essentially consists of direct quotations held together by introductory statements in the author's own words. (Such studies are, unfortunately, still being produced by graduate students in music education.) More experienced investigators often like to use an outline of important points because they can note subject headings and subheadings along with page numbers and enough appropriate annotations to determine which items will be of greatest utility later On. Whatever system of note taking a Person employs, the notations must be gathered with a specific purpose in mind: to relate to the problem being solved, as Mouly avers.14 At the end of each day researchers should alphabetize, by author and title, the cards or sheets they have prepared and place them in their file. In the preliminary examination of a book researchers should first examine the table of contents and the listings in the index; then they should leaf through the book to deterrnine whether or not it is worth scrutinizing further. There obviously is no use in wasting time reading a publication that has few, if any, references listed in the table of contents or the index that are significant to the research. Time could be occupied more efficiently with materials containing pertinent information. It may seem redundant to stress the importance of requisite reading skills since it generally is assumed that by the time one reaches graduate level one already has developed facility in reading rapidly and in distinguishing between what is relevant and unnecessary. Yet sometimes research tyros do not realize that they can save much valuable time by skimrning through an article or book to determine the most significant paragraphs, sentences, and key words. They may then make note of important items and paragraphs that they can reread before going on to new material.

Research in Music Education

104

Burke and Burke, cognizant that note taking frequently is difficult and troublesome because researchers are afraid they will lose some consequential idea as they read, suggest that they have the material reproduced in some type of photocopying process whenever p o s ~ i b l e . ' ~ Some research topics in music education may require using books and articles written in foreign languages. Researchers who lack sufficient foreign-language skill to comprehend what they are reading would do well to consider another topic. To read an article or book almost entirely with a lexicon is a slow, tedious, and frequently unrewarding procedure . In addition to books, articles, and pamphlets, useful information is sometimes found in newspapers. It is helpful to clip articles of significance and place them in a manila envelope or folder. Since newspaper accounts do not always include the year and date of the article, the name of the newspaper and specific month, day, and year should be noted on the margin' or reverse of the clipping. In addition, some researchers list the Page where the material is found.

Obtaining Resource Materials

105

in the library requesting them. A two-week maximum period for using materials should be anticipated, with one renewal possible if there is no demand for them. Several states, through a statewide circulation System, have procedures whereb y books from any library in the state may be obtained for a limited period of time. Rare books and items regarded as irreplaceable if lost rarely, if ever, are obtainable through interlibrary loan. The researcher may then request a copy of noncirculating materials on microfilm or some other type of photocopying process. Microfilm, for instance, is an inexpensive and practical way to procure copies of resource materials that would otherwise be unobtainable. A rnicrofilm reader has become standard equipment today even in the smallest libraries, private or public, so little difficulty should be experienced in locating equipment to read microfilms once they have been received. Microfiche copies of materials, which are cheaper than microfilm, also may be obtained in many libraries having equipment to read them.

Photocopying Materials Obtaining Items from Other Libraries Sometimes materials deemed important for a study are not available locally. It is necessary then to procure them by other means, either through interlibrary loan or by some photocopying process. Interlibrary loan is a relatively easy way to obtain materials from another library. Individuals requesting items on interlibrary loan must know where they may be obtained before they ask their local librarian to procure them (though the librarian will advise on finding locations). A nominal fee to Cover postal charges both ways is normally the only charge. Materials obtained through interlibrary loan often do not circulate and therefore must be used

Various types of photocopying are available to reproduce items either in full-sized or in microtext format. Full-sized copies, either positive or negative, are available through many processes. The photostat, one of the earliest to be used for music, is expensive for extensive copying and often takes considerable time to obtain. More recent processes, such as Xerox, Verifax, Thermofax, and Diazo, have the advantage of providing a copy not only rapidly but also at a lower cost than the photostat. Microforms, both positive and negative, are of four kinds: microfilm, microcard, microfiche, and ultramicrofiche. When a copy of an entire graduate project, rare book, or collection of musical manuscripts is needed, microfilm, either in 16 or 35 mm, provides a very practical and economical way to obtain it. Since microfilm

Research in Music Education

106

has received such wide acceptance in this country, most libraries, as already observed are equipped with at least one microfilm reader. In addition, portable microfilm readers are available for home use. Microcards, of more recent origin than microfilm, require special equipment to enlarge the positive image contained on the 3" X 5" cards. The number of pages on a single card may vary from forty to sixty, depending on the size of the item reproduced. Microcards have begun to receive some acceptance, although not as much as anticipated when they first appeared. They now include not only an American thesis and dissertation publication series but also the Early Music Book Series of significant music books.16 Although they are widely used in Europe the potentialities for microfiche and ultramicrofiche are only now being realized in this country. For example, research studies available through the Educational Resources Information Center of the U.S. Office of Education are obtainable either in hard copy or microfiche. Microfiche may be procured in either positive or negative. A microfiche is essentially a piece of film (3" X 5" or 4" X 6 ) containing frames of microfilm. The 4" X 6" microfiche will accommodate up to eighty pages of textual material. Because microfiche frames are easier to locate and store, the process, although requiring a microfiche reader, has an advantage over microfilm for both the researcher and the librarian. Portable microfiche readers are now standard equipment in many libraries, and a researcher may sign one out for a specified period, often a week. Ultramicrofiche, a piece of film 4" X 6 , can accommodate up to three thousand pages of text. Even though a researcher may have a preference for one microform over another, the availability of items in that microform will determine whether microfilm, microcard, microfiche, or ultramicrofiche will be used.

Obtaining Resource Materials

107

Reference Materials for Music To present an inclusive listing of all references to which the researcher in music education might have recourse would be beyond the purpose of this book. Sources most likely to prove beneficial are suggested in the sections that follow. However, new items appear regularly, and no list can be comprehensive, so researchers need to be continually alert for new materials that might be of assistance. Those seeking specific information in a restricted area of music will find Duckles' Music Reference und Research Materials: An Annotated Bibliography extremely helpfu1.l' Among the English-lanDictionaries, Encycloguage dictionanes, encypedias, Indexes18 clopedias, and indexes most likely to be used are: Baker's Biographical Dictionary of Musicians (Nicolas Slonimsky, ed., 6th ed., 1978); Grove's Dictionary of Music und Musicians (Stanley Sadie, ed., 6th ed., 1969); The Harvard Brief Dictionary of Music (Willi Apel and Ralph T. Daniel, eds., 1960); The Harvard Dictionary of Music (Willi Apel, ed., 2nd ed., 1969); The International Cyclopedia of Music and Musicians (Oscar Thompson, ed., 10th ed., 1975); The New Encyclopedia of Music und Musicians (Waldo S. Pratt, ed., 1936); and The New College Encyclopedia of Music (Jack Westmp and F. L. Harrison, eds., 1976). Among foreign-language dictionaries and encyclopedias for either general or specialized areas are: Die Musik in Geschichte und Gegenwart; allgemeine Enqklopädie der Musik (Friedrich Blume, ed., 1949-); RISM (International Inventory of Musical Sources; 1960-); and Biographie universelle des musiciens er bibliographie gknkrale de la musique (Franqois J. Fktis, ed., 1972). Reference materials in music relating to such diverse subjects

Research in Music Education

108

as musical instruments, musical themes, jazz band, and chamber music, and other areas also are available. Representative examples are: The Complete Encyclopedia of Popular Music und J a u 1900-1950 (Roger D. Kinkle, ed., 1974); An Encyclopedia of the Violin (Alberto Bachmann, ed., 1975); Real-Lexikon der Musikinstrumente, zugleich ein Polyglossar f i r gesamte Instrumenentegebiet; mit 200 Abbildungen (Curt Sachs, 1964); A Dictionary of Musical Themes (Harold Barlow and Samuel Morgenstern, eds., 1975); Band Encyclopedia (Kenneth W. Berger, 1960); Cyclopedic Survey of Chamber Music (Walter W . Cobbett, ed., 2nd ed., 3 vols., 1963); and An Index to Early Music in Selected Anthologies (Ruth B. Hilton, 1977). Often overlooked by musicians General References are such general references as Dictionary of American Biography; Directory of American Scholars; Encyclopaedia Britannica; Encyclopedia of Educational Research; National Union Catalog, with quarterly and amual cumulations; Who's Who in America; Who's Who in American Education; Who's Who in the East (with comparable volumes for the Midwest, West, and South and Southwest); and Who Was Who in America. Periodicals in both music and other Periodicals subject areas often prove to be one of the most valuable sources of information. Researchers can save considerable time by referring to such periodical guides as Education Index (1929-);I9 Guide tu the Performing Arts, 1957-1968; International Repertory of Music Literature (RILM 1967-); Jazz Index ( 1977 -); Music Article Guide (1966-); Music Index (1949-); Music Therapy Index (1976); Vol. 2 is called Music Psychology Index (1978); Popular Music Periodical Index (1973-), which list title and author of article, name of periodical, and date and inclusive pages of article. Most of these guides are up-to-date within a

Obtaining Resource Materials

109

few months. In addition, many joumals indicate on the title Page in what indexes they are covered. An annual index, frequently the last issue of a volume, also appears in most journals. Music journals vary widely in objectives and contents. Some of the representative current American publications in music and music education are: The American Organist (now the A.G.O. und R.C.C. 0. Magazine), The American Music Teacher, American String Teacher, Black Music Research Newsletter, Bulletin of Councilfor Research in Music Education (Illinois), Bulletin of the National Music Council, Choral Journal, The Clarinet, Clavier, College Music Symposium, Computer Music Journal, Contributions to Music Education (Ohio), Current Musicology, Diapason, Ethnomusicology, The Horn Call: Journal of the International Horn Society, Instrumentalist, The International Trumpet Guild Newsletter, The International Trombone Association Journal, Iowa Journal of Research in Music Education, Journal of American Musical Instrument Society, Journal of American Musicological Society, Journal of Band Research, Journal of Church Music, Journal of the International DoubleReed Society, Journal of Jazz Studies, Journal of Music Theory, Journal of Music Therapy, Journal of Research in Music Education, Missouri Journal of Research in Music Education, Music Educators Journal, Music Journal, Musical Analysis, Musical Quarterly, Music Researchers Exchange Newsletter, NACWPI Bulletin, Notes, Perspectives in New Music, Piano Quarterly, PMEA Bulletin of Research in Music Education, T . U.B.A . Newsletter, and Woodwind World-Brass und Percussion. Research reports also appear in many of the journals published by each of the state music educators' associations, other than the research joumals listed. Trade journals, such as Connchord and Selmer Bandwagon, frequently contain articles relating to research of particular interest to instmmentalists. Other sources valuable for research of historical import are periodicals no

Research in Music Education

110

longer published: Educational Music Magazine, Electronic Music Review, and Woodwind Magazine belong in this category. Many colleges and universities prepare bibliographies of graduate research, either cornpleted or in Progress, that are available free of charge or at nominal cost-sorne contain listings in several disciplines; others report only on rnusical studies-as well as reports of sponsored research, both faculty and Student. Music educators engaged in research often will find data in periodicals and journals in other areas. Among those rnost likely to be useful are: American Educational Research Journal, American Journal of Psychology, Educational Technology, Journal of the Acoustical SocieS, of America, Journal of Aesrhetic Education, Journal of Aesthetics and Art Criticism, Journal of Applied Psychology, Journal of Educarional Psychology, Journal of Educational Research, Journal of Experimental Education, Oral H i s t o Association ~ Newsletter, Phi Delta Kappen, Practical Applications of Research (Phi Delta Kappa), Psychological Review, Review of Educational Research, and Scientific American. Local school Systems sometimes print reports of research undertaken by staff rnembers. For exarnple, Research Reports of Los Angeles Ci- School Districts includes some projects in music education. Research in the field of rnusic also is contained in publications frorn various state departments of education, such as the Minnesota National Laboratoq News Bulletin. The ACLS Newslerrer, a publication of the Arnerican Council of Learned Societies, contains reports on postdoctoral research that the organization has funded. Many studies in rnusic are enumerated in its amual report of grants made for research. The rnusic researcher should not overlook such seemingly irrelevant sources for rnusic as the Journal of the American Dental Association. For exarnple, Dr. Heman's article, with bibliography, "Dental Considerations in the Playing of Musical Instruments," should be of interest and value to researchers in instrumental rnusi~.~O

Obtaining Resource Materials

11 1

Two recent publications relating to research in the arts include information that should be useful to music researchers: Murphy and Jones have prepared Research in Arts E d ~ c a t i o n , ~which ' contains brief Summary reports and listings of projects funded by the U.S. Office of Education in aesthetic education, art, rnusic, and theater education, among other areas; Research in the Arts und Aesthetic EducationzZ is a directory of investigators and their fields of inquiry . In addition to educational background and publication history, individuals are listed in Research in Arts und Aesthetic Education according to their professional specialties in the arts and aesthetic education. The following areas of instruction are included in this report: visual arts, music, literature, theater, dance, film rnedia, and interrelated arts. Yearbooks of the Music Yearbooks und Proceedings Educators National Conference and its predecessor, the Music Supervisors National Conference, published annually frorn 1910 to 1940, are especially valuable for the researcher interested in philosophical and curricular trends in music education. Music Education Source Book [I], initially published in 1947; Music in American Education: Source Book Number Two, 1955; and Perspecfives in Music Education: Source Book 111. 1966, are successors to the yearbooks. Proceedings of the Music Teachers National Association, containing addresses and Papers presented at the organization's annual conventions, was published frorn 1876 to 1950. Other important items include Proceedings of the National Association of Music Therapists, Universi~,of Iowa Studies in the Psychology of Music, and the Yearbooks of the ~ ~ 1973 the National Society for the Study of E d u c a t i ~ n .In American Educational Research Association began to publish an annual Review of Research in Education. Additional sources of information are UNESCO publications on rnusic education, yearbooks of the International Society for Music Education, and International Folk Music Council Yearbook.

Research in Music Education

112

Two important kinds of information, often overlooked by music educators, are publications listing the holdings of music libraries and thematic catalogs of incipits. These items are particularly beneficial to someone who is interested in special collections. Thematic Catalog of a Manuscript Collection of Eighteenth-Century Italian Instrumental Music in the U n i v e r s i ~of California, Berkeley, Music Library (by Vincent Duckles and Minnie Elmer, 1963) is an example of a publication by a university library. A representative publication of a municipal public library is the four-volume Catalogue of the Allen A. Brown Collection of Music, a very significant compilation of nineteenth-century Opera and orchestral Scores, put out by the Boston h b l i c Library from 1910 to 1916. Julia Gregory's Catalog of Early Books of Music (1913), yith the 1944 Supplement, represents a valuable*listing of books on music theory published before 1800 that are housed in the Library of Congress in Washington, D.C." In the past, catalogs of thematic incipits have been marked by diversity of format from one library to another, with the result that an investigator sometimes finds the information contained in the card catalog less than satisfactory. The growing acceptance of electronic data processing (EDP) has opened up new potentialities for a more uniform System of cataloging thematic incipits, according to LaRue and Logemann. They caution, however, that EDP is not i n e x p e n s i ~ e . ~ ~ For the researcher who is looking for information on musicalinstrument museums, the Directory of Special Libraries und Information Centers (4th ed., 1977) lists some. Since the vast majority of Dissertations und Theses research in music education at present consists of reports that culminate in a graduate degree, the researcher should consider bibliographical compilations of these items to be a prime source. Special Catalogs

Obtaining Resource Materials

113

William S. Larson compiled two listings of completed master's and doctoral projects entitled Bibliography of Research Studies in Music Education. The first, covering the years 1932 to 1948, appeared in 1949. The second, published as the Volume V, Number 2 (Fall 1957) issue of the Journal of Research in Music Education, includes studies fom 1949 to 1956. Roderick D. Gordon has continued the Larson bibliography of completed research but has resuicted the listing to doctoral studies. Doctoral Dissertation~in Music und Music Education 1957-1963, compiled by Gordon, was published as the Volume XII, Number 1 (Spring 1954) issue of the Journal of Research in Music Education, and supplements have appeared in the Same journal as follows: 1963-67, Volume XVI, Number 2, Summer 1968; 1968-71, Volume XX, Number l, Spring 1972; and 1972-77, Volume XXVI, Number 3, Fall 1978. The International Index of Dissertations und Musicological Works in Progress (edited by Cecil Adkins and Alis Dickinson, 1977) combines Doctoral Dissertations in Musicology und Musicological Works in Progress, and lists 3,200 American doctoral dissertations and works in Progress and about 1,500 from Argentina, Australia, Canada, Europe, Israel, Japan, New Zealand, and South Africa. The leading repository in the United States for foreign dissertations is the Center for Research Libraries, 5721 South Cottage Grove Avenue, Chicago, IL 60637. Students involved with research that deals with personalities or subjects outside the United States should ascertain whether this center can provide some useful materials. Paul Doe has compiled a register to theses on , ~ ~German dissertations in music have been English m u s i ~ and identified by Richard SchaaLZ7 A Selective List of Master's Theses in Musicology, compiled by D. R. de Lerma, was published by the American Musicological Society in 1970 and H. M. Silvey 's compilation of Master's Theses in the Arts und Social Sciences appeared in 1976. Rita M.

Research in Music Education

114

Mead's Docroral Dissertations in American Music: A Classified Bibliography was printed in 1974; and in the same year the Texas Music Educators Association published a Bibliography of Master's Theses und Doctoral Dissertations in Music Completed ut Texas Colleges und Universities, 191 9-1 972. A somewhat dated, but still valuable, regional compilation of master's-degree projects in music education was completed by Borg.28 Research Studies in Education, an amual compilation by Phi Delta Kappa of doctoral dissertations, research reports, and field studies, includes many in music. The foregoing compilations merely list titles of graduate research; investigators who wish to examine the research in detail can procure photographic copies of completed dissertations (either full-sized or rnicroprint) and abstracts. University Microfilms, Inc., began publishing doctoral dissertations on microfilm in 1938. Comprehensive Dissertation InGex ( C D / ) , contains a listing by title only of all dissertations accepted by United States and Canadian institutions from 186 1 to 1972, with amual supplements. For the researcher who is starting to select a topic for investigation, this should be a standard reference. With two or three exceptions, all institutions of higher learning in the United States and Canada that have doctoral programs send their dissertations to University Microfilms. Abstracts of no more than six hundred words are contained in Dissertation Abstracts Inrernational. Music projects are listed under Section A (The Humanities). University Microfilms has issued the recent catalog Doctoral Dissertations Music: 1949-1964, which lists the author, title, order number, and price of microfilm and Xerox copies of approximately seven hundred music projects that have appeared in Disserrarion Abstracts International and its predecessor, Microfilm Absrracrs. DATRIX,a new computerized data-retrieval system for doctoral studies operated by University Microfilms, will be discussed in a later chapter. As noted earlier, some research reports may be obtained on microcard from the Sibley Music

Obtaining Resource Materials

115

Library of the Eastman School of Music. Microcard Edition, Lnc., Washington, D.C., also has a limited number of items on microcard or microfiche of interest to the music researcher. Since November of 1966 Government Publicarions Resources in Education (formerly Research in Education) has been appearing monthly under the sponsorship of the Educational Resources Information Center (ERIC),of the U.S. Office of Education. Currently there are sixteen ERIC clearinghouses, which acquire, select, catalog, abstract, and index the documents listed in Resources in Education (RIE). In 1969 ERIC began publishing CIJE (Current Index to Journals in Educarion), covering over seven hundred major journals that relate to all areas of education. Useful in conjunction with RIE and CIJE is Thesaurus of ERIC Descriptors (7th ed., 1977), which contains all subject headings or descriptors used for retrieval of documents in the ERIC collection. As an example of the types of studies handled by clearinghouses Kemeth S. Goodman's Music Analysis: Applications t o Reading lnstruction, was published in 1973 by the ERIC Clearinghause on Reading and Communications Skills, Urbana, Illinois. In order to make biographical informaNecrology tion complete an investigator needs to discover sources of necrology . To examine all issues of newspapers and periodicals for several years to obtain this kind of information can be a very time-consuming and frequently disappointing pursuit. Yet this undoubtedly is the best way to procure the needed information if it is not available elsewhere. Fortunately for musicians and librarians, each June issue of Nores includes a section on necrology. These accounts list names of the deceased, their nationality and musical distinction, date and place of death, and source or sources of obituary. This writer has made it a continuing practice to clip obituaries

Research in Music Education

116

of musicians and other professionals from newspapers, and then, after indicating date and publication, place them in a Special folder. From time to time the accounts are sorted and clippings appended to appropriate sources containing biographical sketches of these people. Such a system has advantages only if it is regularly pursued over an extended period of time and if the researcher has recourse to a newspaper that has extensive national and worldwide coverage. Other publications, such as the Music Educators Journal and Time, to mention only two with different coverage, also include necrology. Additional sources of necrological information include Biography Index, the New York Times Index for Necrology, and annual supplements to an encyclopedia, such as the Britannica Book of the Year. Maintaining a current chronography of every author listed in the card catalog obviously is not the responsibility of researchers. When data are not forthcoming, however, they must obtain them as best they C&. Studies of a historical or Archives und Central aesthetic nature may make it Repositories necessary for researchers to examine manuscripts, writings, and other artifacts in libraries or archives that are not Open to the general public. An investigator interested in studying the compositions of New England composers Charles C. Perkins (1823-1886) or James C. D. Parker (1828-1916), for exarnple, almost certainly would need to plan a visit to the Harvard Musical Association in boston to examine manuscripts, publications, and other materials of these composers housed there. For one interested in early Moravian music the collections of the Moravian Archives at Bethlehem, Pennsylvania, and Winston-Salem, North Carolina, contain prime source materials. Unlike public libraries, archives frequently are Open only by invitation. Researchers, however, are usually granted the

Obtaining Resource Materials

117

privilege to make visits for research purposes if they follow the protocol governing such matters: (1) ask permission to exarnine a certain collection or collections at a time convenient to the archivist; (2) state purposes and objectives of the visit; and (3) obtain sponsorship from an institution of higher learning or recognized scholar in the field who is interested in the research. Authorities in the field and advisers usually can given excellent advice relative to the most likely repositories containing the information a researcher is seeking. The Midwest Library Center, housing rarely used materials that normally circulate from the libraries of the Big Ten schools, and the New York State Library system are examples of central repositories or circulation systems that enable researchers to procure materials that might otherwise be inaccessible to them. The new developments in information retrieval that are coming to the forefront will make it possible in the near future for scholars to receive within minutes an item they may request. The material will be transmitted by a Computer interchange system and can be reprinted on the spot, as it were, for the use of the one requesting it. The Music Education Historical Research Center, established in 1965 at the University of Maryland in College Park, has become one of the most complete repositories of music-education materials in the United States. Investigators can examine manuscripts, letters, Papers, speeches, textbooks, and other materials, many of them irreplaceable, by the leaders in music education both past and present. Also included are yearbooks, minutes from meetings of music educators, state music-educators association publications, early instmctional method books, and various publications in the field of music education. Several individuals are preparing bibliographies of a composer's works or of articles in various areas of music for use with information-retrieval systems. For example, The International Repertory of Music Literature (RILM), mentioned earlier, a

Research in Music Education

118

computer-indexed international bibliography containing short abstracts of scholarly works about music published since 1967, is under the direct editorship of Barry S. Brook of the City University of New York. One of the most important steps in research, it will be recalled, is that of examining the literature in a chosen field to determine the feasibility of a proposed study. The process seemingly is open-ended because new and significant information appears almost daily. It is important for the investigator either to possess or to develop the proper techniques to make efficient use of the many resources and processes of bibliographic information available. Some kinds of research are almost exclusively dependent on information obtained from published and unpublished sources in libraries, archives, and private collections. Historical research, which best exemplifies this kind of investigation, will be discussed in the next chapter.

Problems for Revie W and Discussion 1. What techniques can make use of the library more effective?

2. Why is the library card catalog an important source of preliminary information? 3 . Compare methods of note taking. Give advantages and disadvantages of each one. 4. Consider the various photocopying processes and indicate which ones are accessible in your local library.

5. Think of a specific music topic and list categories under which reference materials may be found.

Obtaining Resource Materials

1 19

Supplementary Readings ALMACK, JOHNV. Research und Thesis Writing. Boston: Houghton Mifflin, 1930. Chapter 9. ARY,DONALD,LUCYC. JACOBS, and ASGHAR RAZAZVIEH. Introduction to Research in Education, 2nd ed. New York: Holt, Rinehan and Winston, 1979, Chapter 3. Educational Research und Evaluation Methods. ASHER,J. WILLIAM. Boston: Little, Brown, 1976, Chapter 10. and HENRYF. GRAFF.The Modern Researcher, BARZUN,JACQUES, 3rd ed. New York: Harcourt Brace Jovanovich, 1977, Chapters 4, 5, 6. BEST,JOHNW. Research in Education, 3rd ed. Englewood Cliffs, N. J.: Prentice-Hall, 1977, Chapter 3. BRICKMAN, WILLIAMW. Research in Educatiom/ Histon. Norwood, Pa.: Folcroft Library Editions, 1975, Chapters 3 and 6. BURKE.ARVIDJ., and MARYA. BURKE.Documentation in Education, 4th ed., rev. New York: Teachers College Press, 1967, Chapters 2, 3, 5, 10, and 15-22. DUCKLES, VINCENT.Music Reference und Research Materials: An Annotated Bibliography, 3rd ed. New York: The Free Press, 1974. ENGELHART, MAX D. Methods of Educatioml Research. Chicago: Rand McNally, 1972, Chapter 4. Fox, DAVIDJ. The Research Process in Education. New York: Holt, Rinehart and Winston, 1969, Chapter 4. GOOD, CARTERV. Essentials of Educational Research. New York: Appleton-Century-Crofts, 1954, Chapter 3. HOPKINS,CHARLES D. Educational Research: A Structure for Inquiry. Columbus, Ohio: Charles E. Merrill 1976, Chapter 12. JONES,RALPHH., ed. Methods und Techniques of Educational Research. Danville, 111.: Interstate Printers and Publishers, 1973, pp. 35-48. Manual of Music Librariamhip. Ann Arbor: Music Library Association, 1966, Chapters 5 and 10. MOULY,GEORGE J. The Science of Educational Research, 2nd ed. New York: Van Nostrand Reinhold, 1970, Chapter 5.

Research in Music Education

120

SAX, GILBERT. Empirical Foundations of Educational Research. Englewood Cliffs, N.J. : Prentice-Hall, 1968, Chapter 4. SHAFERROBERTJ., ed. A Guide to Historical Method. rev. ed. Homewood 111.: Dorsey Press, 1977, Chapter 4. SMITH,HENRYL. Educational Research, Principles und Practices, Bloomington: Educational Publications, 1944, Chapter 4. TRAVERS, ROBERT M. W. An lntroduction to Educational Research, 4th ed. New York: Macmillan, 1978, Chapter 4. TUCKMAN ,BRUCEW. Conducting Educational Research, 2nd ed. New York: Harcourt Brace Jovanovich, 1978, Chapter 3. VAN DALEN,DEOBOLD B. Understanding Educational Research: An lntroduction. New York: McGraw-Hill, 1979, Chapter 4. WATANABE, RUTH T. lntroduction to Music Research. Englewood Cliffs, N.J.: Prentice-Hall, 1967, Chapters 1-7, 9- 12, 16-17. WHITNEY, FREDERICK L. Thc Elements ofResearch, 3rd ed. Englewood Cliffs, N .J .: Prentice-Hall, 1960, Chapter 4 and Appendix 3. WIERSMA, WILLIAM. Research Methods in Educ-ation: An Introduction, 2nd ed. Itasca, 111.: F. E. Peacock, 1975, Chapter- 12.

5 . Historical Research: A Chronicle of the Past The collection of data pertaining to the past frequently is referred to as history. Vincent notes that "history is a narrative of the actions of human beings in comection with the topic of research, whether it be in politics, law, religion, music, or mechanical arts. ''I Nevins presents a different viewpoint: history is "actually a bndge connecting the past with the presenf and pointing the road to the future. ''= When the data assembled are interpreted the process is known as historiography, a procedure similar to the fourth of Dewey's steps of reflective thinking (accumulation and codification of data) discussed in Chapter 1 of this book. Researchers in music education who use the historical method consequently are ex-

Research in Music Education

122

pected not only to obtain various kinds of data about the past related to their topic, but also to interpret them according to generally accepted techniques of historiography. Historical research, as is tme with all methods or procedures for obtaining data, must employ the scientific method to solve problems in history. Defining an approach that is predicated on the scientific method, Fischer says that history is "a problem-solving discipline" and the historian is someone who "asks an open-ended question about past events and answers it with selected facts which are arranged in the form of an explanatory ~ a r a d i g m . " ~

Purposes of Historical Research Historical research is pursued for a variety of,purposes. Study of the past may serve as the basis for understanding the present or for predicting the future. Although this concept is not used as extensively in music education as it is in the social or political sciences, researchers may study stylistic periods in music, for example, with this objective in mind. The simpler and emotionally restrained Classical period may be better understood when contrasted with the highly chromatic, introspective, and less formal Romantic era. In the same manner, concepts of neoclassicism might become more evident when compared with classicism. A desire to learn more about the life of a significant music educator, contemporary or deceased, or the organization, developrnent, and infiuence of a performing group or a professional organization, presents many opportunities for a researcher who has an inclination in any of these directions. A historical study of the contributions to music and music education of Luther Whiting Mason (1 828- 1896) or Charles H. Fmsworth (1859- 1947) would be exarnples of the former. Representative of the latter might be "The Chicago Symphony and Its Role in the Improve-

Historical Research

123

ment of Musical Standards in Nineteenth-Century United States" or "History of NIMAC and Its lnfluence on Music Education." The quest for information to complete a missing link in the figurative chain of knowledge is a third reason for the student of music education to undertake historical research. Recent investigations into the musical contributions of the Moravians in America is an example of filling a gap in the history of music in America. Undoubtedly the most significant reason for involvement in history today is the Same one given by Thucydides (c.460-c.400 BC),generally acclaimed as the "father " of the modern historical method, who remarked that his purpose was "not to write for immediate applause but for posterity. " This eminent Greek historian also explained very astutely that it was his duty to supplement and compare, as accurately as possible, his own observations of the Peloponnesian War (43 1-404 BC)with the reports of eyewitnesses. Although historical research in music is often regarded as the exclusive province of musicology, there really is little justification for this restrictiveness, as an examination of historical studies in music education over the past twenty years will reveal. Garrett, in defining the function of the musicologist, indicates that the researcher uses scholarly procedures to locate and organize data relating to the evolution of all types of m ~ s i cThese . ~ Same conditions are incumbent upon the music educator engaged in historical research. There is no need, as is commonly done, for investigators who use the scientific method to be rigidly labeled according to some narrow, arbitrary distinction like historian, musician, or educator. Rather, they should be regarded as "research scholars." This is in harmony with the views of Hockett, who considers the usefulness of history as the "intellectual honesty which should be developed by the quest for truth in any field of knowledge, in defiance of all temptations to wander from the strait and narrow path which alone can lead to it. "5

Research in Music Education

124

Historical Research Procedures Historical research is based on certain steps or procedures that, Hockett states, consist of "the gathering of data; the criticism of data; and the presentation of facts, interpretation, and conclusions in readable form."6 The first point is closely related to the sources and kinds of information needed for the study. External and internal criticism, to be discussed later in this chapter, are used respectively to determine the truthfulness and trustworthiness of the data. Hockett's final step, relating to exposition and interpretation of data, already has been referred to as historiography, but will continue to be discussed throughout this chapter. Gottschalk has a slightly different emphasis, asserting that "historical facts have to be (1) selected (2) arranged, (3) emphasized or minimized and (4) placed in some sort of causal sequence. '" Gottschalk's first two steps relate Jo what Hockett calls "gathering of data." His third point bears a relationship to Hockett's "criticism of data" but with positive or negative implications. Finally, Gottschalk's sequence of data placement is closely allied to Hockett's "interpretations and conclusions. " Barzun and Graff present a more succinct statement about research procedures applicable to historical research: the subject or topic "is defined by that group of associated facts und ideas which, when clearly presented in a prescribed amount of space, leave no questions unanswered WITHIN the presentation, even though many questions could be asked OUTSIDE it."* Historical research is sometimes criticized for being too subjective because historiographers are not always in a position to exercise the rigid controls that natural scientists regard as absolutely essential. Historiographers, unlike scientists, normally are concemed with more than an observation of the present. When involved in the preservation of today's chronology, which tomorrow will be history, historiographers must be just as accurate and precise as natural scientists in order to realize the objective of presenting and inter-

Historical Research

125

preting the truth. A music critic, attending the world premiere of a composition by either a renowned or unknown Composer, is on the threshold of history, as it were. The critic's responsibility to the musical world in accurately reporting the performance of this new work is essentially no different from that of the reporter who writes the proceedings of a conference of the world's leaders for dissemination to the general public. Both have an obligation to provide information that is as objective, truthful, and unbiased as is humanly possible. The classic "who, what, where, when, why, " timewom as it may seem, is another way to look at a procedure for collecting historical data. Researchers in history, however, should always be aware that they may never be able to obtain all the information needed and that additional sources may be discovered later that will refute what they believe to be true based on their evidence. The onus is on researchers to verify all data as best they can. Discussing the importance of verification of Statements, Shafer states that a researcher "must make the best judgments he can on every statement that is important to his investigation. In every case the judgment is one of varying degrees of probabilityprobably true, probably accurate, probably untrue, probably inaccurate. "9

Source Materials Source materials used by the historiographer are normally of two kinds, primary and secondary. A primary source of data, as noted in Chapter 3, is "firsthand" information. When data are not original to the one reporting them they become "secondhand" or secondary sources. It is also possible to obtain information from a "tertiary" or other source even further removed from the original. Every effort should be made, though, to obtain original data because the basis of historical research is the pn-

Research in Music Education

126

mary source. According to Mason and Bramble, historiographers recognize the necessity of directing their energies to obtain this kind of information.I0 An investigator preparing the biography of a living person might use such primary source materials as correspondence with and publications by the subject; legal and personal documents, such as contracts, wills, and diaries; newspaper items; photographs and photocopies; concert programs; oral and written reports by contemporaries; and tape recordings of personal interviews. Secondary sources often have value but should be used only when primary sources are untrustworthy or difficult to procure. Brickman asserts, however, that reliable secondary sources are preferable to primary sources that are questionable or incomplete.ll Secondary sources worthy of merit will be adequately documented since they should be based on primary materials. Accounts of a concert written by someone other than the person actually present are secondary, as are specially prepared Summary accounts of the minutes of official meetings. History books and many encyclopedia articles are secondary even though the author may have had recourse to primary materials. Likewise, publications by an individual may be primary sources, but when excerpts from then are quoted by another they become secondary. Biographical accounts of composers frequently include an analysis of their music to indicate how events of their lives may have influenced their creativity. For example, Vallis, in a doctoral study, incorporated biographical information into an aesthetic inquiry that involved an analysis of Brahms's concertos to trace evidences of the Baroque instrumental style in these works.12 Vallis concluded that Brahms was not only a crusader for Baroque music, but that he also incorporated such Baroque elements as counterpoint and the concertato principle into his concertos; the texture and rhythm were characteristically Baroque in many instances.

Historical Research

127

Selection of Topic Although factors in the selection of a topic already have been discussed at considerable length in Chapter 2, five that Brickman considers to be significant before an investigator decides on a historical-research topic are: (1) interest, (2) source materials available, (3) the time factor, (4) relevance to course, and (5) specialized knowledge . I 3 The reader will recognize immediately that these steps differ only slightly in emphasis from those listed in Chapter 2. Researchers in music education should direct their energies toward a topic in which they have an intense interest. A highschool choral director, for instance, might Want to write a history of the high-school choral contest-festival movement in the United States. Preliminary investigation may reveal that source materials are so limited that it would be impractical to continue with the original idea without expanding its scope. Such expansion would obviously require additional time, a factor that might not be feasible in a Situation where a deadline becomes an extemal factor imposed only by the duration of the instructional period. One aspect has not yet been mentioned: Brickman states emphatically that the topic of a research report should bear relevance to the objectives of the research class.14 While there may be exceptions, it is only logical that researchers in music education select a topic that they can develop in relation to music education. One of the principal objectives of a music-education research course is to give students experience in making practical application of the precepts they l e r n there. Taking Brickman's enjoinder more seriously could result in fewer unsuitable musiceducation research proposals. In the realm of specialized knowledge, the researcher who considers a study of the history of music education in Germany, to cite a possibility, should obviously possess both a knowledge of the history of music education in general and an adequate

Research in Music Education

128

reading comprehension of German, the language in which most primary sources would appear. It would be very impractical, as noted earlier, to attempt logical translations with the use of a German-English lexicon exclusively, especially if one were unfamiliar with music-education terminology and philosophical principles.

Oral History The origins of Oral History date back to 1948, when Columbia University professor Allan Nevins conducted an oral-history interview with George McAney, a New York civic leader. The Columbia University Oral History Collection now includes the memoirs of over three thousand people and 375,000 pages of transcript. Oral history is a procedure or method fpr obtaining and preserving historical information in spoken form based on the interview. Data gathered through oral-history interviews are largely intended for future use by historians rather than for the substantiation of hypotheses, as would be the case in experimental or descriptive research. Cutler states that "knowledge about the past passes by word of mouth from generation to generation, and not only must the scholar capture it but he must also try somehow to filter the significant from the inconsequential as well as the truth from the exaggerated."15 The collection of data in oralhistory research is not by means of spontaneous or random interview~,but by the following steps: (1) invitation by the researcher to individuals to participate; (2) preparation for the interview; (3) preinterview visit; (4) the interview session (tape recorded); (5) preparation of verbatim transcript of the interview;

Historical Research

129

(6) submission of transcript to interviewee for corrections and additions, along with legal draft agreement; (7) concluding the legal agreement and contract or statement of gift; and (8) preparation of final copy of interview. Before sending an invitation to prospective interviewees researchers must identify the purposes of the research and the problem to be investigated. If the oral-history project involves working in a remote area, individuals able to provide the desired information must first be identified, and then necessary preparations arranged, such as housing for the researcher, equipment to be used (usually portable battery-operated cassette tape recorder and adequate supply of tapes), and mode of transportation to reach the interviewees. The researcher meets and establishes rapport with the interviewees, indicating purpose of the interview and arranging a definite time and place for the next meeting. Researchers should use a portable tape recorder because they will not have time to write down everything of importance that the Person being interviewed may say. An added advantage is that a tape recording may be replayed as many times as necessary to obtain a literal transcript of the interview. A verbatim transcript allows the interviewee to make additions, corrections, or deletions and to protect both parties from any legal action that might result. Legal documents should be worked out with the interviewer's attomey. Once editorial and legal details have been clarified the final document is submitted for signature, and the contract or statement is made indicating the type of remuneration the interviewee is to receive. The oral-history process culminates in the preparation of the copy to be deposited in some archive or central repository for the benefit of other researchers. Oral-history studies in music education have been infrequent to date, but this should change as researchers become more in-

Research in Music Education

130

terested in sociological and cultural studies that impinge on music. However, recent national and regional meetings of the Music Educators National Conference have included sessions devoted to oral history. One of the most recent oral-history studies in music education was completed by Brobston.16 In the first part of his study Brobston presented the antecedents of gospel music, music used by evangelists in the nineteenth and twentieth centuries, and the current Status of gospel music in the South, including singing conventions, publishers of gospel music, and performers. The second part of his study contained the field research, or oralhistory aspect. This was conducted in twenty-five counties of South Georgia selected at random, from which twenty-three farnily singing groups were interviewed and recorded in actual performance. Brobston concluded that there appears tobe no signifiCant diminution of gospel-singing activity in the geographical area included in his study. The Oral History Association publishes two sources of information that should be helpful to music-education researchers: Oral History Association Newsletter and the annual Oral History Review. Both contain articles, abstracts of studies that have been completed and a listing of recent publications dealing with oral history. One such citation is Cooper's "Popular Songs as Oral History: Teaching Black History Through Contemporary Audio R e s o ~ r c e s . ' "At ~ the time of this writing The Oral History Collection of Columbia University, 1978-1983 was in preparation. It is to be a dictionary-catalog of entries in the Oral History Collection at Columbia University.

Historical Research Data Gathering Musicians involved in historical research will most likely find that their data fall into one or more of the following categories:

Historical Research

131

that of (1) heretofore-unknown information about an individual, group, object, or era; (2) uncovering heretofore-unknown creativity of an individual; (3) uncovenng a more complete or an authentic copy of an individual's work; (4) uncovering an individual's creative efforts that were known to have been written but believed not to be extant; (5) collecting, codifying, or analyzing information of historical import from diverse sources; (6) verifying that documents, statements, or creative works reputedly by an individual are spurious; and (7) rectifying incorrect statements, dates, or information previously accepted as true. One of the most gratifying by-products of historical research is the discovery of heretofore-unknown information. Although such data result from serendipity, opportunities for revelations of this kind are too infrequent in music and music education. The most likely sources of such information are newly discovered manuscripts, personal letters, or documents that previously were unavailable to the general public. Barzun and Graff give an interesting account of the former's discovery of a previously suppressed letter of Berlioz that appeared shortly after the first appearance of his "Rakoczy March" in 1846. The original was not only misdated but also distorted from a letter basically musical in content to one with "political " implications. l 8 Even more significant than Barzun's discovery is the announcement by Richard L. Crocker, professor of music history at the University of California at Berkeley, that the deciphering of a Song on clay tablets pushes the history of music back a thousand years to the second millennium BC. Unearthed by French archaeologists in the 1950s in Ugarit, in what is now Syria, the cuneiform symbols were undecipherable until 1972. The Song is based on the Same heptatonic diatonic scale in use today in Western music. Crocker states that it sounds like a lullaby, hymn, or gentle folk song.19 Discoveries of heretofore-unknown creativity normally are of two kinds: items that have been identified positively as authentic

Research in Music Education

132

and those about which some question remains regarding authenticity. Newspapers and periodicals like the Journal of the American Musicological Socieo and the Musical Quarterly frequently contain accounts of the discovery of lost items. Pauly, for example, reports on some newly discovered manuscripts of Michael Haydn (1737- 1806), a Composer of sacred music and younger brother of Franz Joseph Haydn (1732- 1809). Pauly indicates that as the result of several months of research in libraries in Austria, France, Germany, and Italy he located extensive holdings, previously unknown, of the younger Haydn's manuscripts in the Bavarian State Library in Munich and in monasteries throughout Au~tria.~O The Amencan concert pianist Byron Janis identified a hitherto-lost manuscript copy of Chopin's "Waltz in G Flat Major, " Op. 70, No. 1. Found at the Chateau de Thoiry in Yvelines, France, the manuscript was unearthed in a box marked "old clothes" by the mansion's owner, B u n t Paul de La Panouse, who showed the work to Janis. The connection between Chopin and the Panouse family is that Chopin's friend, French writer Eugene Sue, was in love with the second wife of one of Count Paul's ancestors. Janis, recognizing the importance of the manuscript, had it authenticated by Chopin expert Francois LeSure, of the Societe Francaise de M~sicologie.~' Discovered at the Same time and in the same place was the "E Flat Waltz," Op. 18, the ''Grande Valse Brillante. "22 The discovery a few years ago of a Violin Concerto in C Major, reputedly by Beethoven, is an example of a composition the authenticity of which remains uncertain. Only further investigation will reveal whether this does belong in fact to Beethoven's catalog of works. In addition to locating previously unknown works, another source of personal satisfaction to a researcher is to uncover a more complete or authentic copy of a musical composition or other creative work. This writer, for example, while examining

Historical Research

133

stacks of uncataloged instrumental music in the Moravian Archives in Winston-Salem, North Carolina, in 1949, located a complete set of the Three Trios, Op. 3, of John Antes (174018 1 l), minus the last Page of the violoncello part. John Bland of London published these undated trios, apparently about 1785, because Antes was known to have been in England then. Prior to this wnter's discovery the only known set in existence consisted of a second violin and a violoncello part owned by the Sibley Music Library of the Eastman School of Music in Rochester, New York. A comparison of the Eastman and Winston-Salem copies disclosed exact duplication. Thus, by using the Winston-Salem first and second violin parts and the Eastman violoncello part it was possible to present the first modern performance of Trio in D Minor, Op. 3 , No. 2, the second of these charming Haydnesque trios, at the University of lowa, Iowa City, on May 19, 1950, with violinists Stella Hopper and Joame Dempsey and violoncellist Charles Becker as performing artists. Antes, whose surname is the Greek equivalent of von Blume, was born at Frederick-trop, Montgomery County, Pemsylvania, on March 24, 1740. In accord with prevailing educational philosophy of the Moravians at that time Antes intensively pursued musical instruction in his youth in addition to his other studies. In 1769, after being ordained at Marienborn, Germany, he went to Egypt as a missionary. While recuperating from physical afflictions Antes composed some quartets and other pieces before retuming to Europe in 1782. In Viema, according to Grider, Antes met Haydn, who assisted other musicians in the performance of some of his c o m p o s i t i ~ n sIt . ~was ~ during this period of convalescence in Cairo that the trios were written, as indicated by the inscription of the title page, which states in part "Composti a Grand Cairo da1 Sigre Giovanni A-T-S. Dillettante Americano. Op. 3" (Composed at Grand Cairo by Mr. John A-T-S. American Dilettante. Op. 3.). Quite in keeping with the custom of the time, Antes perplexingly listed himself as A-T-S, a factor that

Research in Music Education

134

resulted in his anonymity until 1940, when Carleton Sprague Smith, then chief of the Music Division of the New York Public Library, revealed the identity of this important Moravian compo~er.~~ When any work is unearthed its importance ultimately may rest on its value when viewed in light of historical perspective. Under ordinary circumstances the discovery of works by someone such as Antes, who is relatively unknown to music educators, probably would go unheralded. In terms of historical perspective, however, the Antes Trios assume a new dimension. They represent, as far as the writer has been able to discern, the earliest extant chamber music written by a native-born A m e r i ~ a n . ~ ~ Another phase of historical research in music education that is bound to bring personal gratification is ferreting out creative efforts known to have been written but believed not to be extant. At the Moravian Archives in Winston-Salem, in the same pile of uncataloged instrumental compositions that c'ontained the Antes Trios, this writer discovered the manuscript of "Parthia IX, " by David Moritz Michael (1 75 1- 1825), generally regarded as the most important nineteenth-century Moravian instrumental composer. Rau and David, in their catalog of American Moravian music, report that "Psrthia IX" was not to be found in Bethiehem . ~ ~and David had indicated that the nor Lititz in P e ~ s y l v a n i aRau collections of Lititz and Bethlehem contained thirteen wind partier~.~'With the finding of "Parthia IX" it may safely be stated that Michael composed at least fourteen partien, all of which exist in manuscript in the Moravian Archives either at Bethlehem, which now contain the former Lititz holdings, or at WinstonSalem.28 Since Bethiehem and Winston-Salem served as ecclesiastical headquarters for the northern and southern divisions of the Moravian Church, respectively, the duplication in manuscript of some compositions in the archives of these two cities may be explained by the constant interchange of personnel.

Historical Research

135

Michael, however, confined his activities to Bethlehem and Nazareth, Pemsylvania. These partien, in accordance with the meaning of the word, were meant to be performed out-of-doors. Grider notes that they usually were performed from a balcony of the home of the Moravian Brethren on Wednesday evenings during the summer for the benefit of the citizens of the c ~ m m u n i t y . ~ ~ In four short technically easy movements, "Parthia IX," except for the second movement, is scored for two clarinets in Bb, two French horns in Bb, and bassoon. The first (Allegro), third (Minuet), and fourth (Allegro) are in the key of B b Major. The second (Andante), in the key of F Major, is scored only for two Bb clarinets and bassoon. The Polish-born violinist Henryk Szeryng, after a seven-year search, finally located a manuscript copy of Concerto No. 3 in E Major, by violin virtuos0 Nicolo Paganini (1782- 1840). It was known that Paganini had written five violin concertos, but prior to Szeryng's rediscovery Concerto No. 3 was the only one that had not been found. Authenticity of the work was verified by French musicologists .30 Keahey, in a doctoral dissertation, analyzed eight tnos by J. S. Bach's youngest son, Johann Christian Bach, that are not found anywhere else. Included also is a thematic index to update Terry 's 1929 biography of the c ~ m p o s e r . ~ ' Collecting, codifying, or analyzing information of historical import from diverse sources, sometimes known as "documentary research," represents another very important aspect of historical research in music education. In a doctoral study exemplifying this type of research Bagdon brought together many sources to present a concise picture of musical activities in Charleston, South Carolina, which in 1735 was the site of the first ballad-opera Performance in America, Flora, Hob in the Well. Bagdon declared that the St. Cecilia Society, founded in 1766, was one of

Research in Music Education

136

the first musical organizations in America, and although musical culture was largely imported from England, teachers and instruments were readily available in C h a r l e ~ t o n . ~ ~ In documentary research investigators assemble data from all available sources, codify them, and present their interpretations. The value of this type of study for both the musical layperson and the professional music educator is obvious. Someone who has neither the time nor the inclination to ferret out answers to questions usually welcomes the opportunity to obtain detailed and comprehensive information from one codified source. It is for this reason that dictionaries and encyclopedias of musical biography , terminology, and other kinds of information are so widely used by music educators and others who are seeking ready references. The annals of historiography are replete with examples of documents, statements, compositions, or items that, although they may have been accepted at one time as authentic, subsequently have been proven to be spurious. An often-quoted example is the account of the fantastic and fanciful Cardiff Giant, unearthed in upstate New York in the late nineteenth century. Good tells of a different type of chicanery involving an individual who received a baccalaureate degree from a music school in 1950 and who, by 1963, through extensive forgery, was able to fabricate a transcript, complete with official registrar's seal, equivalent to two graduate d e g r e e ~ This . ~ ~ writer's evaluation of the authenticity of a string quartet allegedly by Benjamin Franklin and a copy of a composition supposedly by Sidney Lanier will be discussed later in this chapter under "External Criticism. " Numerous examples exist in music as well as in general history for the rectification of statements, dates, or beliefs that previously have been accepted as correct. Brickman, Gottschalk Hockett, and Nevins present several accounts to show that many anecdotes and statements attributed to George Washington, Thomas Jeffer-

Historical Research

137

son, and Abraham Lincoln, among many other distinguished Amencans, have proven to be illusory. The correction of an improperly dated musical example well known to students of music history is cited. Hockett, in reporting on the English round Surner is icurnen in, says: A recent example of misdating due to unskillful use of evidence concems an old musical piece known from its opening words as Sumer is icumen. Nineteenthcentury historians thought it was written about 1240 because that was the date of another piece in the same manuscript. Not until the 1940s was the error discovered when a professor from an American university pointed out that the handwriting of the two pieces was not the same, and that the musical notation of the Sumer did not come into use until long after 1240, making the probable date of Surner about 1310.34 This rectification of date was made by the late Manfred Bukofzer (1910-1955). Another interesting example of misdating is the claim that Boston University, in 1876, was the first institution to grant the Bachelor of Music degree in the United States. Eells, however, relates that a report of U.S. Commissioner of Education for 1873 confirms that the distinction properly belongs to Adrian College in Michigan. He notes that Trustees Minutes of Adrian College, dated June 1873, list Mrs. Mattie B. Pease Lowrie as the recipient of a Bachelor of Music degree in 1873.35Eells also reports that the first honorary degree in music was awarded by Georgetown University to Henry Dielman by Zachary Taylor, the twelfth President of the United States. Eells, quoting Shea, states that this apparently was the only time a President of the United States has been privileged to award honorary d e g r e e ~ . ~ ~ Evaluating results of information has been discussed to some extent in conjunction with the seven kinds of data mentioned

Research in Music Education

138

earlier in this chapter. The historical method, however, is more explicitly concemed with external and internal criticism, or the "how " techniques.

External Criticism or Authenticity Through the process of external (or lower) criticisrn the investigator learns whether or not the subject of scrutiny is authentic. Many aspects constitute the overall meaning of external criticism, but the prime objective is to determine whether or not the item of concern is genuine. Hockett, in defining extemal criticisrn, states: "It examines documents-a comprehensive term which. . . includes not only rnanuscripts but books, pamphlets, maps, and even ancient inscriptions and monuments. "37 Verification of authenticity also is part of the external-criticism procedure. Fischer, insisting that the successful obtcome of factual verification rnay rest on the skill of the researcher, says: "A historian must not merely tell truths, but demonstrate their truthfulness as well. He is judged not simply by his veracity, but by his skill at verification. "38 One of the purposes of extemal criticisrn is to detect forgeries. Shafer points out that in 1648 Kircher published what he alleged was the lost accompanirnent to Pindar's first "Pythian Ode, " a fabrication later proven to be f a l ~ e . ~ ~ Some of the questions that might be asked by a researcher in music education who is engaged in the process of extemal criticism are: (1) Where was the item originally located? Where is it now? (2) 1s this document an original version or a copy? If a copy, where is the original? (3) What is the estimated age of the item? Does it appear to be as old as it should be to be authentic? (4) Are there autographs or other identifications that will rnake the process of verification easier? (5) Is the handwriting (in the case of manuscripts) consistent with other items by the reputed writer? (6) Are there any indications (diaries, newspaper ac-

Historical Research

139

counts, etc.) that such an item may have existed? and (7) 1s there any reason to suspect that this item may be a hoax? In the process of external criticisrn it often becomes necessary to probe auxiliary areas of knowledge. Some of the fields musicians most frequently consult seem to be photography, paleography , sernantics, chronology , genealogy , and cartography . Sometimes they may even find themselves trying to determine the watermark of a document as well as the age and kind of paper used . For answers to the questions posed above through the principles of external criticism, consider an intriguing and beguiling flute and piano composition by Sidney Lanier (1842- 188 1) entitled "Danse des Moucherons. " This writer has in his possession a negative photostat of this short, rhapsodic, chromatic work. In an attempt to locate the original manuscript he checked the holdings of the Lanier Room at Johns Hopkins University in Baltimore, Oglethorpe University in Georgia, and other likely sources without success. In addition, personal correspondence with Lanier's children, each of whom courteously replied, failed to shed any light on rnanuscripts other than those at Johns Hopkins, which, as just noted, does not possess the original to "Danse des Moucherons." Henry W. Lanier, who indicated that his manuscripts had been given to the Lanier collection at Johns Hopkins University, wrote the following regarding his father, a self-taught flutist and poet: "He once said to his wife what I believe to be literally true-that the difficulty with him was not to write down music. The moment he had time and strength, there were songs ready to be born. Several brief sketches may be found in the Lanier Roorn at Johns Hopkins, including his projected "Quartette," "Tuno Religioso" (for two flutes or violas), "La Reve" (three flutes and bass flute), and "Trio for flute, pianoforte and violoncello. " External criticisrn can thus give but inconclusive answers to the first two questions. As to the third, this writer was informed that his copy was reproduced directly from the original manuscript.

Research in Music Education

140

The date 1873 appears after Lanier's name, yet the style of handwriting with regard to tempo and dynamics is inconsistent. The tempo markings in particular appear to be in a more contemporary hand, especially on the last two pages of the six-page composition. In addition, the twelve-line manuscript paper on which the composition was written bears this inscription: "Carl Fischer, New York. Monarch Brand Warranted. " It is tme that Sidney Lanier spent some time in New York from 1870 to 1872, the year in which Carl Fischer arrived in the city. Fischer's first business venture, however, was in musical instruments. It was not until considerably later that the publishing phase of the company began. There are no autographs, holographs, or other identifications on this writer's copy to suggest a negative answer to the fourth question. The composer's name is written in a combination of upper- and lower-case hand lettering. The tempo and dynamic terms, as just noted, appear to be in two different hands, making them difficult to compare with the holograph bf the composer's name and identification of the composition. Starke, in his biographical and critical account of Lanier's life, includes a photostatic copy of the flute part of the second and concluding Page of Lanier's "Gnat Symphony, " which consists of only three and one-half l i n e ~ . ~The ' Page does contain Lanier's account, in a very neat and unhurried handwriting, of the composition, which the Composer describes as a "translation of the sound. " As already related, this writer's copy contains six pages, including the piano part, but the last twelve measures of the flute part in the two versions do not agree. In response to the fifth question, a comparison was made of the item just cited and another one by Lanier, his unaccompanied flute solo "Wind Song, " which Starke states was performed in October of 1874, but apparently was written e ~ l i e rThere . ~ ~ is a strong similarity in handwriting; both appear in a style that shows a lighter, more delicate, and less hurried stroke than that in the copy of "Danse des Moucherons" possessed by this writer.

Historical Research

141

That such an item by this Georgia-bom poet-musician may have existed, the object of the sixth question, has already been established in conjunction with the previous discussion, and the title is listed in Starke's bibliography under "Music by Lanier."43 It also is listed in Centennial Edition of the Works ancl Letters of Sidney L ~ n i e r . ~ ~ In answer to the seventh question, it would appear that the authenticity of this copy of "Danse des Moucherons" must be Open to serious question because of the discrepancies that exist, although the composition is listed by both Starke and Graham. Perhaps additional research will result in more convincing data to suggest otherwise. Hockett pinpoints the universal dilemma faced by historians who must pass judgment on a document, especially if there are some questions about its veracity, prudently noting that additional evidence eventually may appear to alter previously accepted i d e a ~ . ~ ~ To cite another example of the application of principles of extemal criticism, this writer became intrigued by an item that appeared in a now-defunct New York City newspaper a few years ago. The article, a dispatch from Lynchburg, Virginia, stated in part that a previously unknown quartet for strings by Benjamin Franklin was soon to have its initial performance in Philadelphia, largely through the efforts of a woman who was an associate professor of French at a local woman's college. The professor's attention was called to this composition by the owner of a Parisian bookstore in which she was browsing. The article stated that the quartet was discovered by an eminent musicologist in a pile of forgotten works, although their location was not d i s ~ l o s e d . ~ ~ Several factors should be noted relative to this quartet. The original manuscript, in tablature, was reputed by the transcriber to be housed in the Bibliotheque Nationale in Paris. In correspondence with this writer, however, music librarian E. Lebeau indicated: "La Bibliotheque nationale, et la Bibliotheque du Conservatoire qui en fait partie, ne possedent ni quatuor ni aucune

Research in Music Education

142

Oeuvre musicale de Banjarnin Franklin. "47 [The national library and the library of the Conservatory which is a Part of it, possess neither the quartet nor any other musical composition of Benjamin Franklin. ] The real motive behind the use of tablature would be interesting to leam. In the preface to his transcription, which was published in 1945, the transcriber reports that the original is in the handwriting of a professional copyist of the late eighteenth century, who otherwise remains unidentified. Benjamin Franklin served as United States Ambassador to France from 1776 to 1785, and it is known that he attended concerts in Paris, including performances of chamber music at the salon of Madame Helvetius d'Auteui1. It does seem rather Strange, though, that tablature would be used for a string quartet so late in the eighteenth century, since this type of writing, except for guitar and other fretted instruments, had largely been replaced by the conventionai system of notation currently in use. Also unexplained is the rationale for placing each of the four instruments (three violins and violoncello) in scordatura. The use of this system of mistuning by Heinrich Biber (1644-1704) is well known, but the simple, uninteresting, Single melody line employed for each instrument in the alleged Franklin quartet suggests no need for scordatura, which normally was used with multiple stops. By using scordatura, however, instrumentalists can play each of the parts in the Franklin quartet entirely on Open strings, another curious circumstance. The quartet is in five short movements. The first bears no subtitle, but appears to be in a march tempo. Other movements are: menuetto, capriccio, menuetto, and siciliano. The capriccio and siciliano are basically tripartite while the others are in two parts structurally. Unusual too, is the inclusion of five movements in a quartet of this period in musical history, as is the use of three violins and violoncello. Information from the Franklin Institute of Philadelphia, where the first contemporary performance of the quartet was scheduled

Historical Research

143

was to the effect that the work had not yet been presented. Furthermore, all attempts to communicate with the discoverer and the transcriber of the work proved to be fmitless. These repeated efforts continued for approximately two years, until the death of the transcriber. A review of a performance of the alleged Franklin quartet appeared in a New York newspaper, but no additional information was provided to suggest that the work was genuine, except that the version examined by this writer had been further edited and corrected by yet another i n d i v i d ~ a l . ~ ~ Finally, no definite proof has been advanced to show that Franklin actually knew enough about the technique of musical composition to enable him to compose a quartet, or any other musical work for that matter. True, he often is credited with the invention of the armonica, or musical glasses, a distinction he disclaimed. Franklin, however, admittedly did make many improvements on the instrument, whose sweet and pleasant tinkling Sounds were produced by moistened fingers massaging rotating tumblers of different frequencies. The armonica, or musical glasSes, so intrigued Mozart and Beethoven that they composed several pieces for the instrument. In view of the rather consistent pattem of negative evidence, or lack of evidence, it must be conceded that this work unquestionably is spurious. The Franklin quartet appears to be a musical joke, the product of an anonymous jester. It always is possible, though, that more positive data may be forthcoming in the future that could result in a more exact evaluation.

Interna1 Criticism or Credibility Had the evidence regarding the Lanier and Franklin compositions been more positive, investigation of other factors, known as internal (or higher) criticism, could have proceeded. There are many occasions when this additional step is necessary. Although

Research in Music Education

144

extemal criticism may establish that a document, item, or statement is authentic, there may be inaccuracies or inconsistencies within, as Nevins reports: "While it is sometimes troublesome to establish the authenticity of a historical source, that is usually an easy matter compared with making certain of the entire integrity of a source. "49 The purpose of intemal criticism, according to Gottschalk, is to establish whether data are "at all credible, and if so, to what extent. Experienced researchers frequently engage in external and intemal criticism simultaneously, in which they will use information from one to assist in the implementation of the other. Some historiographers divide intemal criticism into positive and negative phases. Making a distinction between them, Brickman states that positive criticism tries to ascertain the true meaning of Statements, whereas negative criticism concems a researcher's rationale for discrediting them as evidenced by the partiality or inefficiency displayed by the ~ r i t C r . ~ l The investigator in music education who is dealing with internal criticism must answer the questions: (1) Is the document consistent stylistically with others by the writer? Are there major inconsistencies? (2) Are there any indications that the writer's reporting was inaccurate? (3) Does the writer actually mean what he or she says? (4) Could this work have been written by someone else in the style of the individual? and (5) 1s there any evidence that the writer is biased or prejudiced? lt generally is conceded that artists do change their manner of expressing themselves stylistically due to the natural process of artistic growth and development. For example, students of music literature are regularly concemed with comparing various characteristics of a composer's stylistic periods. One of the most obvious examples is the three distinctive epochs in the creative life of Ludwig van Beethoven (1 770- 1827); the first ending about 1802, the second approximately in 1815, and the third in 1827. Although Composers may show stylistic changes during their lifetime, the musical idioms that they favor usually persist

Historical Research

145

throughout their creative life. Characteristic idioms serve as guides for the researcher when attempting to ascertain whether the composition in question is consistent with others by the Same composer. Major inconsistencies may result from the composer's deliberately altering an accustomed style by using different idioms. On the other hand, the incongruent features also suggest that these works are indeed by two different individuals. In the instance of a composer like Arnold Schoenberg (1874- 195l), a change in both stylistic characteristics and musical idioms is apparent in the transition from the postromantic style of the Verklärte Nacht to dodecaphonic Pierrot Lunaire. Occasionally inaccuracies will be suspected in a document. Barzun and Graff describe the incorrect dating of a letter by Berlioz to his publisher indicating that he was forwarding the Table of Contents to a book. Berlioz dated his communication "Thursday June 23"; Barzun and Graff point out that the year must have been 1852, in which case June 23 fell on Wednesday. They also remind their readers that it was not unusual for Berlioz to mistake the day of the ~ e e k . ~ ~ A third question that might be raised in regard to internal criticism is whether or not the writer actually meant what he or she said. Although some confusion understandably could come from writers assuming that their readers comprehend and agree with their definition and use of certain terminology, misunderstandings are often more than merely problems of semantics. Witness persons who use the word "cornett" when "comet" is intended. Despite the difference of only a single t the instruments are vastly dissimilar in nature. The former refers to an obsolete fifteenth- and sixteenth-century instrument, normally made of wood, containing six finger holes and played with a cup-shaped mouthpiece. The contemporary comet, a three-valved instrument constructed of metal, is similar to the trumpet, but shorter. Furthermore, the cup-shaped mouthpiece of the cornet is larger and deeper than that of the comet. The next question, more applicable to music than other disci-

Research in Music Education

146

plines, is concemed with compositions that may have been written by an imposter in the style of another Person. Especially suspect are "newly discovered " works by such well-known composers as Haydn and Mozart. Reasons for this type of deception vary, but one of the most common is to enable someone who is unheralded to capitalize on the name of an esteemed Composer. On the other hand, there are numerous examples in which musicians, in good faith and with no attempt at deception, have completed works left unfinished by composers at the time of their death. A represe ntative example is Mozart 's Requiem, which was finished by his protege and intimate friend Franz X. Sussmayer. Despite the best of intentions, it is difficult to be completely unbiased or unprejudiced . Some writers, of Course, deliberately present a unilateral point of view, as evinced by some early textbooks on the history of music in the United States. The treatise by Frederic L. Ritter, Music in America, for instance, is generally regarded as strongly prejudiced in the'direction of German Romanticism. In a general sense, the realm of aesthetic judgment is characterized by personal predilection and preference, just as is the choice of an automobile. It likely would be more difficult for a critic who prefers the piano music of Liszt to that of Chopin to present an unbiased account when reviewing the works of both composers than it would be for one who has no preference. Likewise, two concert artists do not perfonn the Same composition in precisely the Same manner. Gottschalk, aware of these differences in interpretation, comrnents " . . . perhaps it is desirable that they should. Musicians are, after a fashion, merely historians interpreting past achievements of a specialized nature. "53 Historical research can provide many challenging opportunities for a music educator to make significant contributions to human knowledge. However, it should be remembered that history is more than reporting facts. Nevins succinctly avers that "a balance must be maintained between history as a body of facts, based

Historical Research

147

solidly upon research, and history as an exercise in analysis, interpretation, and generalization. "54 By using principles of historiography the researcher in music can employ the scientific method to obtain and evaluate facts objectively. Experimental research, which undoubtedly uses the scientific method more consistently and effectively than any other, will be the subject of examination in the next chapter.

Problems for Review and Discussion 1. For what reasons does a music-education researcher pursue historical research? 2. Differentiate between primary and secondary sources and give some specific examples of each. How do you determine whether primary or secondary sources should be used? 3 . Into what categories might data for historical research in music education fall? Give specific examples. 4. Discuss the place of external criticism in historical research. What specific questions might a music-education researcher ask when involved in extemal criticism?

5. What function does internal criticism serve in historical research? Differentiate between the negative and positive phases of intemal criticism. What specific questions might be asked by the music-education researcher engaged in applying the principles of intemal criticism? 6. What is the purpose of oral history? What procedures should be followed to complete an oral-history research project?

Research in Music Education

148

Supplernentary Readings ALMACK,JOHNC. Research und Thesis Writing. Boston: Houghton Mifflin, 1930, Chapter 7. ARY,DONALD, LUCYC. JACOBS, and ASCHAR RAZAVIEH. Introduction to Research in Education, 2nd ed. New York: Holt, Rinehart and Winston, 1979, Chapter 11. BARZUN, JACQUES, and HENRYF. GRAFF.The Modern Researcher, 3rd ed. New York: Harcourt Brace Jovanovich, 1977, Chapters 5-10. BEST,JOHNW. Research in Education, 3rd ed. Englewood Cliffs, N.J.: Prentice-Hall 1977, Chapter 10. BORG,WALTERR . , and MEREDITH D. GALL.Educational Research: An Introduction, 3rd ed. New York: Longman, 1979, Chapter I 1. BRICKMAN, WILLIAMW. Research in Edilcational History. Norwood, Pa.: Folcroft Library Editions. 1975, Chapters 2, 4, and 5. COHEN.MORRIS R . , and ERNEST NAGEL.An Introduction to Logic und Scientific Method. New York: Harcourt Brace, 1934, Chapter 17. R . , and N. KENNETHLAFLEUR. A Guide to Educational COOK,DAVID Research, 2nd ed. Boston: Allyn and Bacon, 1975, Chapter 3. FISCHER, DAVIDH. Historian's Fallacies: Toward a Logic of Historical Thought. New York: Harper and Row, 1970, Chapters 3, 5, 9-1 1. Fox, DAVIDJ. The Research Process in Education. New York: Holt, Rinehart and Winston, 1969, Chapter 14. GARRETT,ALLEN M. An Introduction to Research in Music. Washington, D.C.: The Catholic University of America Press, 1958, Chapters 9 and 10. G ~ D CARTER , V. Essentials of Educational Research. New York: Appleton-Century -Crofts, 1966, Chapter 4. GOTTSCHALK, LOUIS. Understanding History, 2nd ed. New York: Alfred A. Knopf, 1969, Chapters 3, 5-7. HILLWAY, TYRUS.Introditction toResearch, 2nd ed. Boston: Houghton Mifflin, 1964, Chapter 11. HOCKETT, HOMERC. The Critical Method in Historical Research und Writing. New York: Macmillan, 1955, pp. 13-70. HOPKINS, CHARLES D. Educational Research: A Structure for Inquiry. Columbus, Ohio: Charles E. Merrill 1976, Chapter 8.

Historical Research

149

JONES,RALPHH., ed. Methods und Techniques of Educational Research. Danville, 111.: Interstate Printers and Publishers, 1973, pp. 151-180. MOULY,GEORGE J. The Science of Educational Research, 2nd ed. New York: Van Nostrand Reinhold, 1970, Chapter 8. NEVINS, ALLAN.The Gateway to History, rev. ed. Garden City, N.Y .: Anchor Books, 1962. PHELPS,ROGERP. "The Mendelssohn Quintet Club: A Milestone in American Music Education," Journal of Research in Music Education, Spring 1960, 8,1:39-44. SHAFER,ROBERTJ., ed. A Guide to Historical Method, rev. ed. Homewood, 111.: Dorsey Press, 1977, Chapters 5 and 6. SMITH,HENRYL. Educational Research, Principles und Practices. Bloomington: Educational Publications, 1944, Chapter 7. TRAVERS, ROBERTM. W. An Introduction to Educational Research, 4th ed. New York: Macmillan, 1978, Chapter 14. B. Understanding Educational Research: An VAN DALEN,DEOBOLD Introduction. New York: McGraw-Hill 1979, Chapter 11. VINCENT,JOHN M. Aids to Historical Research. New York: Appleton-Century-Crofts, 1934, Chapters 2, 5, and 12. WHITNEY,FREDERICK L. The Elements of Research, 3rd ed. Englewood Cliffs, N.J .: Prentice-Hall, 1950, Chapter 8. WILLIAMSON, JOHNB., DAVIDA . KARP,and JOHNR. DALPHIN.The Research Craft: An Introduction to Social Science Methods. Boston: Little, Brown, 1977, Chapter 10. WISE, JOHN E., ROBERTB. NORDBERG,and DONALDJ. REITZ. Methods of Research in Education. Boston: D. C. Heath, 1967, Chapter 4.

Experimental Research

6 . Experimental Research: Purview of the Future Experimental research usually is regarded as the most objective of the five methods of obtaining data discussed in this book. Because of the rigid controls exercised in experimental research it is possible to duplicate this kind of study and receive virtually the Same results, sornething that is not as easy to accomplish in historical and descriptive research or philosophical and aesthetic inquiry. Experimental laboratory, or empirical research is sometimes regarded as "a glimpse into the future" because it can show what is possible under carefully controlled situations. Fox suggests that this orientation to the future is possible because the researcher is trying to evaluate something new.' Another term associated with this type of investigation is "behavioral research. " Even though researchers often have an idea

151

of what their results will be, with better controls possible in experimental studies the results usually are more definitive than in other kinds of research. Research in the laboratory has long been regarded as the "method of science. " It is from the laboratory that scientific concepts were adapted and applied to educational, including music-educational research. Experimental studies in music education and in education, however, usually attempt to sirnulate empirically, as closely as possible, laboratory conditions in the classroom or in the field with individuals or groups rather than with the process followed in the laboratory with individuals. The term "empirical" refers to information obtained through some type of observation, "metempirical, " through reflective thinking. Precise controls characterize the experimental method. For this reason music researchers have turned increasingly to the behavioral or experimental process, sometimes to the exclusion of other procedures, for obtaining practical answers to questions, as . ~ though an empirical process this writer has clearly ~ t a t e dEven will be used to carry out the research, logic and common sense are necessary to organize an experimental research study, as Beveridge points out.' Experimental research can be simple or complicated, depending on the topic. Keppel's Statement is simple, precise, and direct: "Basically, the experimental method consists of the contrast between two treatment conditions. The subjects in both of these conditions are treated identically, except for one feature that is different. "4 The treatment condition may consist of one or more independent and one or more dependent variables. A variable is defined as "anything that can change." for example, test Scores, grades, or experimental conditions. The independent variable, also known as stimulus, treatment, experimental, or antecedent variable, is the factor that causes a change in something. It is the "manipulation of the independent variable" by a researcher, according to Gay, that "differentiates all experimental research

Research in Music Education

152

from the other methods of research. " 5 The dependent variable, also known as response or criterion variable, is observed to See what happens to it as a result of manipulation of the independent variable. It rnay be defined as the result of the presence or absence of the independent variable. Rating scales and direct observational schedules are examples of dependent variables. This process of determining the relationship between two treatment conditions, the effect of the independent on the dependent variable, is sometimes referred to as "cause and effect." The designations of "independent" and "dependent" for a particular item are not fixed, however. What rnay be an independent variable in one situation rnay be a dependent variable in another. The designation of one or the other depends on how they are used. In 1923 W. A. McCall published a treatise laying out procedures for conducting experimental research in the field of education. Sir Ronald A. Fisher's Statistical Methods for Research Workers (1925) provided further direction f6r the researcher through the introduction of a procedure that was to become known as randomization, according to Campbell and S t a n l e ~ . ~ McCall's concepts were further refined by experimental psychologist Robert S. Woodworth and others in the 1930s and 1940s. One of the most widely used methods for experimental study in music education involves making comparisons between a control group and one or more experimental groups. The control group maintains a "status quo" situation: investigators continue their usual procedure in working with it. Any changes or alterations of method or content take place in the experimental groups, or in the ones that can effect change in the dependent variable. In medical research some new drug rnay be given to the experimental group while a placebo, a harmless substance that looks identical to the drug being tested, rnay be administered to the control group. The placebo helps achieve more uniform controls for the investigator. This is possible because the subjects do not know whether they are receiving the ingredient being tested

Experimental Research

153

or the placebo. Thus they have no reason for reacting abnormally to the experiment, as they might do if they were aware of the content of the medication. Application of the principle of the placebo sometimes is possible in music-education research, but its use generally is quite limited. There are occasions in educational research, which includes music, where psychological factors can differentially affect experimental and control groups when the control group receives no treatment, according to Borg and Gall, who state that a ' 'placebo treatment " could be given to both experimental and control groups to keep psychological conditions similar for both groups. For example, if the experimental group receives a new approach to foreign-language instruction, the control group might experience some background music not presented before . The music would not affect t he experimental group, and the control group would receive a "placebo treatment, " thus creating a similar psychological effect.' The concept of the single variable is applicable in almost all situations in experimental research to test a hypothesis. Implicit in the use of the single-variable principle is that all other factors in the experiment remain constant. On the other hand, manipulation of the independent variable results in observable changed conditions, or what is sometimes called a "cause and effect" relationship. Frequently the terms "independent" and "dependent" variable refer to cause and effect. Here is an oversimplified example of hypothetical experimental study. The objective is to learn what effect drill on a certain rhythmic Pattern will have on a student's ability to recognize it in musical context. The drill is the independent variable, and the reading skill that results is the dependent variable. Another experiment might be conducted to determine what effect listening to dodecaphonic music has on a theory student who is writing conventional four-part-harmony exercises. In this example dodecaphonic music is the independent variable, and the dependent variable is the student's harmony achievement. Not all differences observable in the dependent variable rnay

Research in Music Education

154

be attributed to the experimental factor or the independent variable. Some dissimilarities will result from errors due to conditions that cannot be regulated. Lindquist has labeled these kinds of errors as Types S, G, and R.s Type S, or sampling, errors may take place as a result of random sampling. Those under Type G relate to extrinsic factors that may affect one group but not another, hence are known as group errors. When an experiment is repeated, Type R, or replication, errors may be expected because of uncontrollable variations, such as changes in mood of the groups being tested that are present in one administration of the experiment but not in others. Investigators, aware that such errors may occur, should take this into consideration and organize their study to include procedures that measure these deviations statistically . One of the criticisms of experimental research has been that it often contains an element of artificiality, especjally if it is conducted in a laboratory Situation. However, a well-designed experiment that has rigid controls will minimize artificiality, according ~ or behavioral, research, like any to W i e r ~ m a .Experimental other type, needs to be designed to obtain answers to questions in the easiest and most direct way possible. Research conducted by the experimental method essentially is of two kinds: laboratory and field studies. Laboratory, also known as basic, research is perhaps best exemplified in music education by doctoral or postgraduate experimental projects that are concerned directly with finding realistic solutions to basic problems and only indirectly with financial benefits. In a doctoral study Harris investigated the effect of an intonation training program upon the intonation achievement of junior and senior high-school wind instrumentalists. Using his own Instrumental Test of lntonational Skills (ITIS)as a pre- and posttest, he found that the experimental-group students scored significantly higher at the . O l level on the posttest than on the pretest.1° Colnot, also

Experimental Research

155

in a doctoral study, used text and audio-cueing to test the effects of these media on cognitive understandings, attitude toward teaching, and rates and ratios of teacher-approved responses of three groups of university instrumental student teachers. The audio-cueing System resulted in increased academic approval rates and ratios.ll The focus of Petzold's five-year longitudinal postdoctoral study was aural perception of children's musical development. Examining rhythm, timbre, and harmony in a series of pilot studies, Petzold found that the ability to imitate musical ideas was not a measure of the understanding of these ideas.12 Spohn's postdoctoral research tested and evaluated a clinical type of instructional program based on individual differences. He used three procedures: item analysis of student worksheets and test Papers, discriminative analysis to serve as bases for autoinstructional phases, and an experiment using freshman music students enrolled in a music-fundamentals Course at Ohio State University in 1966- 1967. Results showed that individual and group learning problems could be ascertained.I3 In addition to doctoral and postdoctoral research, there are some studies in music therapy that fit the category of behavioral or experimental research, such as the one by Nucci. Before receiving the Ph.D., Nucci had earned an M.D. She used selected types of music to ascertain the concomitants of verbalization of depressed and schizophrenic patients as observed in individual psychotherapeutic settings. Nucci, a psychiatrist, with the assistance of three raters, whose qualifications had been validated by the Lucas and Ludwick Mood Scale, observed eighteen patients (nine schizophrenic and nine depressed) in individual psychotherapeutic settings under three conditions: melancholy music, no music, and stimulating music. Results indicated that depressed patients verbalized more with melancholy music, but stimulating music decreased the number of verbalizations. schizophrenic patients verbalized more with stimulating music,

Research in Music Education

156

while melancholy music decreased the number of verbalizations .l4 In contrast are field studies or classroom research. Normally employing classes, either intact or divided into sections, depending on what the investigator is looking for and how the experiment is organized, this type of research is more common in music education than is laboratory research, especially if the latter deals with only one individual or a very small group of individuals. Many illustrious individuals have distinguished themselves in experimental research. Some of these are Herbart and Pestalozzi (education), Binet, Terman, and Otis (psychology), Harvey (blood circulation), Faraday (electricity), and Seashore, Kwalwasser, and Wing (psychology of music). Music therapy is now gaining acceptance by both musicians and physicians, so studies in this area, largely of an experimental nature, are beginning to make an impact on the field. About fifty years ago Hyde studied the influence of music on pulse rate, systolic and diastolic blood pressure, and velocity of blood flow.15 More recently Dreher, in a doctoral study, investigated the relationship between the verbal reports of subjects and galvanic skin responses to different kinds of music.16 Sears's doctoral study related to the kinds of music that can be effective in changing levels of tension in college students. He found that music is slightly more effective in causing tension changes in nonmusicians than in musicians, and in females than in males.l7

Purpose of Experimental Research Experimental research shows what can result under carefully controlled conditions. Observations of these precisely regulated factors, as already noted, enable a researcher to discem what commonly is referred to as a "cause and effect" relationship. Also known as "causation, " this concept, widely used in experi-

Experimental Research

157

mental research, is based on a set of canons forrnulated by the nineteenth-century English philosopher John Stuart Mill:

( I ) method of agreement (2) method of differences, (3) joint method of agreement and difference, (4) method of residues, and ( 5 ) method of concomitant variations.Is The first canon, the best known and simplest of the five, states that if the factors relating to a certain event have one characteristic in common, that characteristic variable is the "cause" of the effect. According to the second canon, when two or more sets of circumstances are identical except for one factor and when a given effect is noticed only when that factor is present, that factor is the cause of the effect. Both the first and second canons are methods of elimination. The methods of canons 1 and 2 are combined in the procedure for the third canon. In this canon the common factor first is isolated and then is withdrawn to determine if the effect is dependent upon the presence of the common factor. A more complex canon, the fourth, is the method of residue, in which causes are determined by process of elimination. When definite factors are known to cause certain parts of a given effect, the other aspects of the total effect must be due to the remaining factors. The final canon states that when two factors change consistently or when they vary together, the variations in one are caused by the other, or both are affected by the same cause .19 In educational research, including music, solutions to some of the most pressing problems do not lend themselves to the experimental method because of the rigid controls imposed by this type of research. For instance, one of the most vexatious problems confronting music teachers is that of scheduling adequate rehearsal time for performing groups. The descriptive method might be selected to survey what policies regarding rehearsal schedules are

Research in Music Education

158

in effect in various schools. On the other hand, if the music educator were concemed with how to make rehearsals more effective, two procedural methods might be experimentally contrasted and the results compared statistically to determine which one appears to be more effective.

Kinds of Experimental Research As already noted, two types of research characterize the experimental method: the laboratory, or basic, and the classroom, or field, study. In the laboratory experiment, the researcher rigidly controls all independent variables and eliminates extraneous influences that may affect the dependent variable as much as possible. Another strength of the laboratory experiment is that the researcher can use random assignment of subjects and thus control l McGrath one or more of the independent variables. ~ u n k e and state that in a laboratory experiment the investigator organizes the research to focus attention on selected behavioral procedures and the conditions related to these p r ~ c e s s e s . ~ ~ The laboratory study, which usually Centers on the observation of one person or a small group of persons, may be in a laboratory setting, in a location familiar to the subject or subjects (often abbreviated Ss), or in an especially arranged Situation outside the laboratory but simulating laboratory conditions. Some researchers have criticized the laboratory experiment because of the artificiality of the research setting. It usually is organized to exclude distractions that may be present in a research study. However, this is a strength rather than a weakness; the more distracting factors are controlled or eliminated, the stronger the research. Kerlinger states that an even greater weakness of the laboratory experiment is the lack of strength of the independent variables. This is because the laboratory experiment is created for Special purposes .21 Kerlinger gives three related purposes of laboratory research:

Experimental Research

159

(1) to find out the relations that exist under uncontaminated conditions, (2) to test predictions derived from this and other research, and (3) to develop new theories based on the refinement of existing theories and h y p ~ t h e s e s . ~ ~ A good example of laboratory research conducted with only one individual is the longitudinal, or extended, study by Revesz of the Hungarian child prodigy Erwin Nyiregyhazi, which is discussed in Chapter 7. Also of interest to musicians should be the ongoing experimentation related to improving musical instruments by highly specialized personnel in laboratories of the Conn Corporation and other musical-instrument manufacturers. Classroom, or group, research sometimes is referred to as field experimentation or research. In field studies it is often difficult to control all independent variables, and for this reason they sometimes are labeled as "quasi-experimental. " Campbell and Stanley comment: There are many natural social settings in which the research person can introduce something like experimental design into his scheduling of data collection procedures (e.g., the whom and to whom of measurement), even though he lacks full control over the scheduling of experimental stimuli (the when and to whom of exposure and the ability to randomize exposures) which makes a true experiment p o ~ s i b l e . ~ ~

Despite their limitations, field experiments are especially suited to kinds of complex social and educational problems to be addressed by social psychologists, sociologists, and educational researchers. In addition to the difficulty of controlling the independent variables, another weakness is randomization, because subjects in the control and experimental groups do not receive the Same treatment. The attitude of the researcher and lack of precision are two other weaknesses of field experiments mentioned by Kerlinger.24Kerlinger's point is that the researchers' egos may be deflated before they even embark on their field study because

Research in Music Education

160

they have had little or no encouragement from school administrators or from their Peers. His reference to the lack of precision pertains to the difficulty in accurately measuring the effect of the independent on the dependent variables because extraneous variables, such as intenuptions and other noises, are present in classroom and field situations even though every effort has been made to control them. Classroom experimentation is useful where evaluation is needed for a new teaching method, curriculum innovation, or teaching procedure. Kerlinger points out that a compelling reason for conducting field studies is that they offer the opportunity to observe a larger number of subjects simultaneously in the classroom, a setting more natural than the laboratory and involving less student distraction. The incidence of what is commonly called the "Hawthorne Effect " may therefore be less pronounced. When subjects are aware that they are participating in an experiment or recognize that they are receiving special attention, some of the improvement (or regression) may be attributed to the Hawthorne Effect, which derives its name from studies by industrial psychologists in the late 1920s at the Chicago plant of the Western Electric Corporation. Although difficult to do, the Hawthorne Effect can be controlled by introducing a second control group that represents a "systematic intervention and interaction on the part of the experimenter with the subjects, " according ~ instance a new procedure is introduced that to T ~ c k m a nIn. ~this is not expected to have special effects related to treatment or intervention variables being evaluated. Borg and Ga11 refer to this intervention as the "John Henry Effect," from the legend of the worker who pitted his strength and skill in driving spikes in railroad ties against a machine. The John Henry Effect is particularly evident when teaching methods are being compared and the investigator who is working with the control group feels threatened and makes every effort to prove that the traditional method of teaching is equivalent to or better than the experimental m e t h ~ d . ~ '

Experimental Research

161

Sometimes unreliable data are obtained because there is a confounding of variables. Variables are considered to be confounded when additional factors at variance with the independent variables are introduced, or occur. For example, testing control and experimental groups at different times of the day, or using different researchers at different times during the experiment, could cause a confounding of variables. Random assignment of subjects to groups will help to reduce the incidence of confounding of variables. Examples of experimental research conducted by music educators using the classroom method follow. Zimmerman and Sechrest, in a postdoctoral study, designed and administered a series of five experiments to 679 elementary and junior highschool students over a two-year period to determine the relevance of Piaget's concept of conservation to musical leaming. Musical tasks consisted of stimulus Patterns and variations of these patterns. The authors concluded that an early acquaintance with basic music structures and vocabulary is very important and that perception must be kept clear of biasing aspects of music by using musical v a r i a t i o n ~ .Dzik, ~ ~ in a doctoral study, compared two methods of instruction to ascertain whether there would be any change in attitudes toward opera of selected high-school general-music students. Eight general-music classes (four experimental, four control) were taught Bizet's opera Carmen. Students in the experimental classes participated in creativedramatics activities; the control classes only were told the story as they listened to the music. Results indicated that a positive attitude toward opera was developed with the experimental clas~es.~~

Concepts of Experimental Research The keystone of experimental research is the formulation and testing of hypotheses. A hypothesis is a question for which an

Research in Music Education

162

answer is being sought, or-as stated in Chapter 3-a shrewd guess as to the outcome of the research. A hypothesis is based on facts that are already known and on conditions that are expected to exist, but can be c o n f h e d only by testing the relationships that may be evident between the known and the presumed. According to Travers,30 hypotheses should be: (1) Clear und succinct. Use technical terms to avoid such generalities as "good leaming conditions. " (2) Testable. A music researcher who hypothesizes that string performers have a keener sensitivity to pitch than woodwind players would be hard pressed to obtain valid data to accept or reject this hypothesis. (3) Indicative of relationshi~s between variables. When relationships between variables are not clearly stated, the hypothesis is not testable. For a researcher to hypothesize that politically conservative music teachers are more effective teachers than those who hold liberal views is absurd. The hypothesis is not testable because "conservative " and "liberal " are not variables that can be delineated clearly . ( 4 ) Limited in scope. If a research tyro were to state that music instruction in Illinois is more effective than that in New York State, the investigator would be stating a hypothesis that is obviously global in scope, and one that is not readily testable. (5) Consistent with known f ~ c t s . Since a hypothesis is based on information al-

Experimental Research

163

ready at hand, it would be incorrect to hypothesize that all trombone players can read the alt0 clef simply because some music of the Baroque and Classical periods contains trombone parts written in the alt0 clef. Bass trombone players, for example, have little need to l e r n the alto clef because the part they play is almost exclusively in the bass clef. (6) Stated in simple terms. Related to Travers's first criterion, simplicity is not related to significance of hypotheses. That a tone produced on a wind instmment will continue until it is intempted by some intervening force (tongue, termination of breath) is a hypothesis that could be easily tested. (7) Feasible to test within a reasonable time. The passage of time will affect the outComes of a study. Musical maturity of a child who studies privately from the fourth to the seventh grade is a case in point. The technical facility of the fourth-grade Student cannot be effectively compared with the Student's skill as a seventh-grader because the passage of time changes the variables.

Testing Hypotheses Hypotheses must not only be logical but testable. Keppel gives one of the clearest Statements on hypothesis testing, indicating that the process includes the following steps: (1) formulating the

Research in Music Education

164

Experimental Research

research hypothesis, (2) drawing a sample or samples from the population, (3) determining the experimental design, (4) manipulating the independent variables, (5) comparing the results statistically, and (6) accepting or rejecting the hypotheses based on the results of (5).31 After formulating the hypotheses, researchers need to determine the population or universe from which they will draw the sample. They will also need to plan the sampling process carefully. Meaningful results rnay not be forthcoming, even though the hypotheses and sample are valid, if the experimental design used to test the hypotheses is incorrect. Such a design could result in rival or competing hypotheses, in which the answers obtained are due to confounding factors. When the appropriate design for the experiment has been determined the experiment can proceed, and the investigator is then in a position to manipulate or introduce the experimental variable (the "X" factor), to See whether or not the independent variable does effect difference in the dependent variable or variables. After the experiment has been completed the results will be compared using the statistical procedures that best seem to treat the data, taking into consideration what was tested. The final step is a judgmental one. In some instances all hypotheses rnay be rejected or accepted, or some may be accepted and others rejected, based on the comparison of results with appropriate statistical tables.

I

1

Selection of Subjects Three procedures are normally used to select subjects for a research study: randomization, matching, and ranking. Randomization, which means that every person in a group has an equal chance of being chosen, consists of two separate functions: random sampling and random assignment. The first relates to drawing a random sample from a population. A popula-

3

I

1

1' 1 '

j

165

tion is defined as all the potential individuals in a universe or Situation. For example, there rnay be a population of four thousand in the universe of flutists from which a random sample could be drawn, but only a certain number of them actually will be used in a study. In randomization it is assumed that any normal group will possess characteristics that rnay be found in a comparable group. Seventh-graders in School A with a certain background who have had specific training in a musical area rnay be assumed to be comparable to seventh-graders in School B under similar conditions. A random sample rnay be drawn in one of several ways: drawing names out of a hat; flipping a coin; throwing dice; or using a fable of random numbers, which is found in most statistics books. After the random sarnple has been drawn from the population the Ss are randomly assigned to experimental and control groups to achieve equivalency of groups. The experimental group is the one that will receive the treatment, or "X," factor while the conditions for the control group remain the Same, with no attempt to manipulate the independent variable with the control group. Matching is predicated on assigning Ss to either the experimental or control group based on Scores obtained from some instrument used to measure characteristics pertinent to the study. An individual who has a certain score is paired with another who has exactly the Same score. Ss who cannot be paired are eliminated from the study. Suppose ten ninth-graders received the following raw scores on the pitch section of the Seashore Measures of Musical Talents: 48, 48, 47, 45, 45, 45, 40, 38, 38, 38. The first subject with the score of 48 would be assigned to the experimental group and the other 48 to the control group. The subject with the score of 47 would not be used in the study because no other person has an identical score. The first two 45s constitute a pair and would be assigned to the experimental and control groups, respectively. The subject with the score of 40 and the one with the third 38 are excluded because there is no match

Research in Music Education

166

for them. Although a high degree of comparability is obtained through matching, the necessity of eliminating Ss is a serious weakness. Ranking provides for the assignment of Ss alternately to the experimental and control groups, even though it may not always be possible to pair the scores. In the previous example 48 and 48 would be paired and would be assigned to the experimental and control groups, respectively. The subject with 47 would go to the experimental group and the first 45 to the control group. The next two 45s would be paired in experimental and control groups, respectively; the 40 would go to the former and the first 38 to the latter. Finally, the last two 38s would be paired and placed in experimental and control groups, respectively. Although the scores of all Ss are used, there will be instances when the Ss paired will not have matching scores, thus making the ranking process weak if there is a concern for compaIjng the growth of paired individuals who do not have matchings scores. Before matching or ranking can take place a pretest could be given to Ss to be included in the research. A pretest is some type of measurement to determine the performance of Ss prior to the begiming of the experiment. One procedure would be to introduce the "X" factor to the experimental group, but not to the control group, and then at the end of the experiment to administer a posttest to both groups to ascertain whether the "X" factor actually did make a difference. The same instmment, or equivalent forms of the same test, should be used for the pretest and posttest. Equivalent forms will be discussed in the next chapter.

Sampling Since it usually is not possible to investigate the characteristics of the entire population or universe relevant to a study , it is necessary to use a procedure known as sampling.The sample, however,

Experimental Research

167

must be representative of the population for data thus derived to be valid. When data on all factors under consideration can be obtained for every subject in a population it is unnecessary to use sampling, but this is extremely rare. It might be both practical and possible to procure information from all living individuals who either have taught or currently are teaching harp in a state with a population of 5 million. To obtain a response from every living Person who either has taught or currently is teaching any kind of instrumental music in that state, however, might be both impossible and impractical from the standpoint of time and money. A sampling of instrumental music teachers should be undertaken. In determining the specific nature of a sample, three factors need to be kept in mind: (1) the definition of the population, (2) where this population is located, and (3) delimitation of this representative sample. In defining the population an investigator might restrict the study to teachers of wind instruments and would then need to determine who all of the wind-instniment teachers are and where they are located. Listings of music teachers could be obtained from such diverse sources as telephone books, city directories, rnembership lists of music educators, union membership lists, rosters of school personnel. and the services-offered section of want-ad columns in newspapers. It might then be decided to delimit the representative sample to private teachers of wind instruments. The group subsequently selected would be representative of the entire population of wind-instrument teachers. Any inferences drawn about wind-instmment teachers would be based on information given by the sample group. The procedures used to select the sample are crucial to the outcomes of the research. Research findings must be generalizable beyond the sample. For exarnple, to compare the musical achievement of sixty fourth-grade students in suburban schools A, B, C, and D with the musical achievement of sixty fourthgraders in inner-city schools D, E, F, and G would be appropriate as long as generalizations for either category of fourth-graders

Research in Music Education

168

were made to the other groups in a comparable population. The difference between characteristics of the population from which the sample is drawn is called sampling error. Sampling error is directly proportional to the size of the sample: the larger the sarnple, the smaller the error, the smaller the sample, the larger the error. Random, systematic, stratified, and cluster sampling are the four procedures for sampling most often used, according to Borg and Another possibility is volunteer sampling, although the Ss usually are biased. In a random sample each individual in a population has an equal chance of being chosen or selected. Sampling errors can be minimized through inferential statistical procedures. A random sample may be drawn in one of several ways. One of the most widely used is a table of random numbers, found in most statistics books. The tables usually consist of a series of five-digit numbers generated by a Computer. A Segment of a hypothetical table of random numbers is given in Figure 5. If 109 Ss are to be selected from a list of 875 students the researcher could start at any place on the table and move horizontally, vertically, or diagonally. For instance, if the researcher who

Figure 5 Random-Numbers Table

Experimental Research

169

wishes to move vertically started in Column 4, Row 3, the Student listed with the last three digits 680 would be chosen first, 093 would be next, 844 would be next, and so On. In column 4, Row 5, 942 is omitted because it is larger than 875, the total number of Ss to be selected. When only three digits are used the first two of the five-digit configuration are ignored. In systematic sampling a sample is again drawn from a population, and the Ss are placed in random order on a list. The researcher then decides on a formula to be used to select Ss-for example, every eighth individual on the list. To start, any subject up to the eighth one on the list can be used, then every eighth subject after that will be selected. Systematic sampling is a somewhat easier procedure to use than random sampling because each subject is not chosen independently, as would be the case with random sampling. In stratified sampling the sample is chosen in such a way that all subgroups in the population will be represented in the sample in proportion to their representation in the population. When all subgroups are represented comparisons between various subgroups can be made. For example, a researcher might Iike to find out if a correlation exists between musical-literacy Scores on Level 6 of the Zowa Tests of Music Literacy (ZTML) and three levels of intelligence of eighth-graden.33 Since there are generally recognized differences between boys and girls in any school System, both in general intelligence and in musical literacy, it would be necessaty to select a stratified sarnple that includes the following levels of intelligence: highest-range boys, highestrange girls, middle-range boys, middle-range girls, lowest-range boys, lowest-range girls. Based on the proportion of students in each level, a random sample would be drawn that is proportional to the number of students represented in each group of the population. Cluster sampling is used when it is more convenient to select a group of persons representative of the population rather than pro-

Research in Music Education

170

portional subgroups, as would be the case in stratified sampling. Cluster sampling deals with groups that occur naturally. For instance, a researcher who wishes to study the musical characteristics of high-school juniors who play the French horn might use cluster sampling when lists from which junior French-horn players could be selected are not available. The researcher could divide the school districts in a state into certain discrete areas, or clusters-thirty-two, for example. After numbering each area, the researcher would draw the schools to be sampled in each area at random, and the junior French-hom players at random from the sample. Volunteer samples are used extensively in education and music-education research because it usually is possible to obtain more information from these Ss than would be the case with random sampling. One major difficulty with volunteer samples is the bias of Ss. Those who may have very positive or negative positions on a topic usually are the ones who volunteer to Seme as research Ss. Because of recent legislation, ethical and legal constraints pose an additional burden for the researcher who deals with volunteer samples. The Family Educational Rights and Privacy Act of 1974, also known as the Buckley Amendment, and the National Research Act of 1974 are of most concem to the researcher. Ethical considerations are those to which professional organizations, such as the Music Educators National Conference, the American Educational Research Association, and the American Psychological Association, expect an investigator to subscribe.

The Pilot Study Before researchers launch a full-scale investigation they should conduct a pilot study to determine whether the intemention hypothesized between independent and dependent variables will materialize in a situation comparable to the one proposed for the

Experimental Research

171

actual study. Runkel and McGrath point out that a pilot study is used to confirm one's guess about characteristics of a given situat i ~ n . 'A~ smaller number of Ss is used in a pilot study as compared to the actual research. For example, if it is proposed to use four hundred Ss in the actual research, forty might be sufficient for a pilot study. The Ss used in the pilot study must not be included in the actual study, although both samples should come from a comparable population.

Designs for Conducting Experimental Research Campbell and Stanley have developed a schema for conducting experiments that has become a Standard for the educational res e a r ~ h e r . 'Included ~ are three tables that list sources of invalidity for the sixteen designs discussed. The authors divide the designs into three categories: preexperimental, designs 1-3; true experimental, designs 4- 6; and quasi-experimental, designs 7- 16. In addition to including the notation and discussion of each of the sixteen designs, factors that could jeopardize internal and external validity are presented. Tests of internal validity determine whether the independent actually effects a change on the dependent variable. To put it another way, are the results obtained due to the effect of the independent variable? Intemal validity is the sine qua non for ideal research designs. according to Campbell and S t a n l e ~ . 'The ~ amount of generalization the research results will have on similar groups is the focus of extemal validity. However, external is more difficult to achieve than intemal validity . According to Campbell and Stanley, there are eight factors or extraneous variables that could jeopardize intemal validity if they are not controlled: history, maturation, testing, instrumentation, statistical regression, selection biases, experimental mortality, and selection-maturation intera~tion.~'

Research in Music Education

172

History, the first factor, is concemed with what has happened between the pretest and the posttest. Has there been some unusual event, such as the Jonestown, Guyana, Massacre of 1978, that could change the reaction of Ss to the experiment? In an experiment to measure degree of like or dislike for classical music, the subjects rnay have attended a live performance of Beethoven's Symphony Na. 6 in F Major between the pretest and application of the independent variable or treatment. Although it rnay not have been related to the experiment, the Beethoven performance could have been a source of internal invalidity because if Ss had never attended a symphony concert before, it could have been historically significant for them. Since Ss are growing older and more mature due to the Passage of time. maturation is the second factor that needs to be kept in mind. The time of day the experiment is conducted is an important maturation variable. When Ss are hungry or {ired they rnay not perform as well on an experiment. If music students have heard a composition before, their musical taste could have matured and their reaction to repeated hearings of it (maturity) could result in a source of intemal invalidity. A third factor that rnay jeopardize intemal validity is the practice effect of having taken a pretest. This could affect the results of the second testing or posttest, which is administered at the close of the experiment. Intemal invalidity is possible even when equivalent forms of a test are administered, for example Forms A and B of the Musical Memory section of the Drake Musical Aptitude Test. 38 Instrumentation, the fourth factor, relates to changes that take place because different observers or scoring procedures are used between the pretest and posttest. Occasionally the calibration of a measuring instrument changes, and this also could jeopardize intemal validity. On an ear-training test, if the procedure for recording the correct answer were changed from oral to written, a source of internal invalidity could exist. .

Experimental Research

173

The fifth factor that rnay jeopardize internal validity, statistical regression, occurs when extreme scores of Ss in a sample are used. An experimenter rnay believe that results would be more significant statistically if the Ss were drawn from the highest and lowest quartiles, or the upper 25 percent and the lower 25 percent, respectively, omitting the Ss who rank in the middle 50 percent on a pretest. Statistical regression means that scores in the upper quartile tend to regress downward toward the mean or average on the posttest and those in the lower quartile tend to regress upward to the mean. Selection bias, the sixth factor, occurs when an experimenter has previously worked with some of the Ss included in the sample but not with others. If an investigator has already had certain students in a music class prior to the experiment, these rnay unintentionally be given preferential treatment over others simply because the researcher knows their capabilities. Volunteer Ss also constitute a selection bias. When Ss are no longer available for the posttest, a seventh factor could jeopardize intemal validity: experimental rnortality. In this instance Ss have moved or are no longer available for other reasons. Only scores of Ss who take both pretest and posttest rnay be included in the evaluation and reporting of research results. High music-test scores for students who are no longer emoiled in the experimental music class, for instance, rnay not be used to prop up less satisfactory scores of those still remaining in the program. Selection-maturation interaction, the eighth factor, occurs when the effects of a design that has not been the best one for the expenment confound results that might have been expected from the independent variable. The question facing the researcher is whether the results obtained are the djrect result of the independent variable or due to the confounding of another variable. The music researcher who employed one specific music test to determine the results of Computer-Assisted Instruction (CAI) might ~

~

Research in Music Education

174

wonder if another test should have been used if the results were not what were expected. Campbell and Stanley list four factors that could jeopardize external validity: reactive effects of testing, interaction effects of selection biases and the independent variable, reactive effects of experimental arrangements, and multiple-treatment interference. A pretest could alter the responsiveness of a subject to the experiment so that results for Ss who have a pretest might be different from those in the Same experiment who do not have a pretest. The second source of extemal invalidity is the interaction effects of selection biases und the independent variable. When Ss are selected in a biased manner the results obtained from manipulation of the independent variable cannot be generalized with any degree of confidence. Reacrive arra ngements, the third factor, refcr to the artificiality of the setting for much experimental research in education because students are aware that they are participating in an experiment . Multiple-treatment interference, the fourth factor, occurs when several treatments are given to the Same Ss and the effects of each preceding treatment are not completely obliterated. It is usually more difficult to control external than intemal validity. Ideally, a research design should make provisions to control both. Since this is not often possible, however, a design that possesses strong controls for internal, but few for extemal validity , should not necessarily be regarded as weak. The designs that follow use certain notational symbols: X the independent variable

0 the observation, or result of the independent on the dependent variable R

randornization to achieve pretest quality of groups .

Experimental Research

175

Parallel rows separated by straight lines represent groups equated by randomization; parallel rows separated by dashes represent groups not equated by randomization. An X and an 0 in one row apply to the SE at different times; an X and an 0 vertical to one another represent simultaneous conditions. Campbell and Stanley regard three designs as preexperimental:39 ( I ) one-shot case study, (2) one-group pretest-posttest design, and (3) static-group comparison. All these designs are considered weak and unacceptable for solid research because they contain so many sources of internal and extemal invalidity. Of the three, the strongest is (3), static-group comparison Notation for these preexperimental designs is: (1) One-shot Case Study

(2) One-Group Pretest-Posttest Design (3) Static-Group Comparison -X- -00 . Campbell and Stanley regard three designs as true experimental designs because all possible factors jeopardizing intemal validity can be controlled, as well as some of those of external validity. The true experimental designs with their notation are: (4) Pretest-Posttest Control-Group Design

(5) Solomon Four-Group Design

Research in Music Education

176

Experimental Research

177

(10) Nonequivalent Control Group Design

0 x- - -0-

- --

(6) Posttest-only Control-Group Design

0 0 (1 1) Counterbalanced Design (sometimes called Latin Square)

Time 1 Time 2 Time 3 Time 4 Designs 7 through 16 represent quasi-experimental designs. These are frequently associated with descriptive research because they are used in social Settings or in situations that call for observation of Ss. In quasi-experimental situations the experimenter introduces data-collection procedures but often cannot exercise full control over the independent variables of the "X" factors. The quasi-experimental designs do not control sources of internal invalidity as well as do true experimental designs; however, some of them-12, 13, 15, and 16-control sources of ixternal invalidity better than designs 4, 5, and 6 under the true experimentaldesign designation. Quasi-experimental designs listed by Campbell and Stanley, with the notation for each, are: (7) Time Series

(8) Equivalent Time-Samples Design

The subzeros in this instance, and those that follow in subsequent designs, represent hazards to external validity because of the difficulty of spacing when "X " is repeated.

(9) Equivalent Materials Design

Group A XI 0

X, 0

XB0

X4 0

GroupB

X,O

X40

XiO

X30

Group C

X3 0

X, 0

X, 0

X, 0

Group D X, 0

XJ 0

X, 0

X, 0 .

In this design each "X," or treatment, occurs only once in each row and in each column. (12) Separate-Sample Pretest-Posttest Design

R X 0. The parenthetical (X) is a treatment that is not relevant to the experiment . (13) Separate-Sample Pretest-Posttest Control Group Design R 0 (X)

(14) Multiple Time-Series Design

Research in Music Education

178

(15) Recurrent Institutional Cycle Design: A "Patched-Up " Design ClassA

XO,

----------_--

Class B, Class B,

R R

0 2

X X

0 3

o4 o5X .

- - - - - - - - - _ - - _--_

Class C

This complicated design, in which the dependent variable, or X, is introduced with different groups (classes) at different times, is not as practical to use as are others listed above and should be used only as a last resort. (16) Regression-Discontinuity Analysis This design was developed to replace ex post facto designs. No notation is given for this design, but it is based on'the premise of awards that are made to students. Students who received awards were compared with those who had other achievements. The purpose of the regression-discontinuity analysis example given by Campbell and Stanley was to See whether or not the awards actually made a d i f f e r e n ~ e The . ~ ~ inference made is that those receiving the awards would have scored higher even without the awards when compared with those who did not receive awards.

Experimental Research

179

consist of two kinds of music instruction, CAI and traditional, and two levels of music aptitude, high and low, as illustrated in Figure 6. In Figure 6 Group A, high-musical-aptitude students, receive CAI, high-scoring Group B receive traditional instruction, lowsconng Group C students receive CAI, and low-scoring Group D receive traditional instruction. High-scoring students would be assigned randomly to Groups A and B and low-sconng students to Groups C and D. According to Gay, in a 2 X 2 design there is one variable that is manipulated and one that is not.42 In Figure 6 miisical aptitude is the control variable, or the one that is not manipulated.

Ex Post Facto Research Ex post facto research is regarded as a specific type of research by Kerlinger, and is frequently encountered in studies by sociologists or behavioral s ~ i e n t i s t s Literally .~~ meaning "after

Figure 6 2 X 2 Factorial Design

Traditional

Factorial Designs Factonal designs enable an investigator to measure the effect of two or more independent variables simultaneously and to examine their interaction with one another. According to Gay, the term "factorial" means :hat the design involves several factors. Factorial designs can be simple or complex. For example, the simplest, a 2 X 2 factonal design, involves four groups and could

Scores

Group C

Group D

Research in Music Education

180

Experimental Research

the fact, " this procedure is quasi-experimental in scope, because experimental procedures are simulated and cause and effect are determined after they actually have taken place, thereby introducing a certain amount of subjectivity into the study. A serious weakness of expost facto research is the difficulty of controlling the independent variable since it must be reconstmcted after the circumstance has occurred. Kerlinger lists two additional weaknesses of ex post facto research: inability to randomize groups, . ~ ~because such and difficulty in interpreting data p r ~ p e r l y Yet variables as personality, musical or other kinds of aptitude, home background, and parental control are not manipulable, ex post facto research is important in certain kinds of studies.

Basic Statistical Concepts No textbook on research methods can Cover adequately all the statistical concepts necessary for data-interpretation purposes. However. a basic knowledge of certain statistical concepts is important so that all researchers will be able to understand the research reports they read. Courses and textbooks specifically devoted to statistics should serve the reader as the basis for a complete understanding of the subject. There are several reasons for Reasons for Using using statistical procedures. Statistics Through the use of statistics data may be: (1) codified, (2) analyzed, and (3) interpreted. lmplicit in the first designation are both the compilation and codification of data. Scores or results of testing or making comparisons are of little value unless they can be collected in one place and classified in a manner useful to the investigator. In an obvious example, when test scores of fifty trumpet players are to be related to the scores of fifty clarinetists, it is important that two separate listings be main-

t

i; I/

$ 1 11

I I

~i

*T ' .I 1

I I

t ?

18 1

tained, one giving trumpet and the other clarinet data. Then these findings can be compared effectively. Once data have been collected they need to be analyzed to determine possible relationships. Comparisons might be made between individuals within each group or between the groups as a whole. The interpretation of data is concerned with an attempt to find out whether differences are due to the effects of the experimental treatment (the independent variable or X) or may be accounted for by chance alone. Statistics are either descriptive Kinds of Statistics or inferential. Descriptive, "parasometimes called metric," statistics describe or summarize discrete bits of information that comprise the results of an experiment or study. In descriptive statistics the relationships among scores obtained from comparison units of reasonably equal size may be Seen graphically and objectively. Inferential, or "nonparametric," statistics refer to making valid inferences from samples to whole populations. Descriptive statistics numerically "describe " a particular group or groups; inferential statistics "infer" what may be assumed to be representative of a large group or population. Inferential statistics also specify the degree of error that may be anticipated when these inferences are made. The most important types of descriptive statistics are measures of central tendency, measures of variance or variability, measures of relationship, and standard scores or measures of relative position. Inferential statistics to be discussed are: standard error of the mean, tests of significance, degrees of freedom, the t test, analysis of variance, analysis of covariance, and chi Square. The most frequently used and most important measure of Measures of Central central tendency is the mean, Tendency or arithmetic average of Scores.

Research in Music Education

182

Experimental Research

Computed by adding all the scores and dividing the total by the number of scores, the mean is identified by X or M. The formula for finding the mean is

X

=

I

Cx N

in which

X

=

the mean or arithmetic average of the scores

C = upper case Greek letter sigma, meaning

"the sum

of " X = each score or measurement in the array N = number of measurements. Assume that scores of 35, 37, 40, 43, and 45 are obtained on a music-aptitude lest Then CX = 200, divided b y N = 5, resulting in X = 40. The median, designated by Md or Mdn, the second most useful measure of central tendency, represents the middle point in a distribution of scores. Half the scores in a distribution occur above and half below the median. In the comparison of scores just given for the mean, 40 is both the median and the mean; thus, the distribution is symmetrical. Mode, a third measure of central tendency, is indicated by the symbol Mo. This term denotes the most prevalent interval in a distribution of scores. For example, in a tally of the scores 4, 8, 4, 6, 4, 1, and 7, the mode is 4 because it appears more frequently than any other number. A distribution containing two noncontiguous intervals where scores are concentrated is called bimodal, asevidencedbythesefigures: 3 , 6 , 7 , 4 , 3 , 2 , 9 , 7 , 1 , 3 , 8 , 5 , and7. In this instance 3 and 7 are the numbers that make this distribution bimodal. The mode usually is the least useful of the three measures of central tendency just discussed because it may fluctuate from one Set of data to the next.

W

i

183

Variability , or variance, relates to the amount of deviation or Measures of V a r i a b i l i ~ spread of scores from the mean. Gay points out that it is possible for two sets of data that are very different to have identical means or medians, but there still is need to determine the variance of individual scores from the mean.45The three most frequently used measures of variability are range, quartile deviation, and standard deviation. Range is the difference between the highest and lowest scores in a distribution. When scores are close together the range is small; when scores are far apart the range is large. Range is frequently used to give quickly an unofficial estimate of range of variability . Quartile deviation, or semi-interquartile range, provides a method by which variability can be determined with reasonable accuracy. One-quarter, or 25 percent, of the scores fall below Q, , or the first quartile; three-quarters, or 75 percent, of the scores are below Q3, or the third quartile, and so On. The semi-interquartile range contains about 50 percent of the scores and is considered to be the difference between 75 percent and 25 percent, or between Q3 and Q , , divided by two. Because scores of both the highest and lowest are eliminated, there is less likelihood that extreme scores will influence the quartile deviation. When the quartile deviation is small the scores are close together, and when they are scattered the quartile deviation is large. Standard deviation is the most useful and frequently used measure of variability, or variance. After a determination is made by subtracting each score from the mean, the differences are squared. Then the squared differences are added and the total is divided by the number of scores resulting in the amount of variance from the mean. The Square root of the variance is called standard deviation. Standard deviation is indicated by the formula ( T = *

N

Research in Music Education

184

in which cr = lower-case Greek letter sigma, meaning standard deviation ZX' = sum of squared deviations from the mean N = number of measurements or scores. For example, 3, 15, 6, 10, 5, and 9 total 48, divided by 6 or N, results in X = 8. Subtracting each Score from X results in 5, -7, 2, -2, 3 , and - 1. When the deviations from X are squared they become 25, 49, 4, 4, 9, and 1, which signify an arithmetic total of 92. When 92 is divided by N (or 6) the variance is 15.33. The U , determined by taking the Square root of 15.33, is 3.92. Calculation of U is a necessary first step for many advanced statistical techniques. For statistical purposes it is assumed that the scores of all individuals may be plotted along a normal probability curve, also known as a "bell-shaped" or "Gaussian" curve, as illustrated in Figure 7.46In this hypothetical example it is assumed that the

Figure 7 Normal Probability Curve

Experimental Research

185

mean or most common or critical value is 0. Deviations from this central point are listed as either positive or negative and the total for all components of the curve equals approximately 100 perCent. It will be observed that there are as many scores to the right as there are to the left of 0 in Figure 7. This means that 50 percent of the scores are above and 50 percent are below the mean. Also, the median and mode would be the Same. In Figure 7 approximately 68 percent of all scores in a normal curve will fall between + 1.0 and - 1.0 standard deviations, and 95 percent of all scores will be found between +2.0 and -2.0 standard deviations (68 percent + 27 percent). Between +3.0 and -3.0 standard deviations will be approximately 99 percent of all scores in a normal distribution. The tails of a normal curve are asymptotic, that is, they never touch the horizontal line or abscissa, but always approach it. When there are more extreme scores at one end of the array than the other, the distribution is skewed and "the mean will always be in the direction of the greater number of extreme scores," report Borg and Ga11.47 In Figure 8, where the tail is skewed to the right, the mean is to the right of the median. When a curve is skewed negatively to the left, as in Figure 9, the mean is to the left of the median. Relationship is concerned with Measures of Relationship determining whether or not a correlation exists between two or Correlation or more variables, and if one does exist, the amount of that correlation or relationship. Correlation may vary from a positive one of 1.00 to the negative - 1.00. The more the two scores agree, the more positive the correlation; conversely, the more they disagree, the more negative the correlation. The two most frequently used techniques for determining correlation are the Pearson product-moment correlation, or r, and the Spearman rank-difference, or rho, designated by the lower-case

+

Research in Music Education Figure 8 Positively Skewed Curve

186

Experimental Research

187

Greek letter p. Product-moment correlation, which shows deviation from 0 in a frequency distribution, may be computed by the formula

product-moment correlation coefficient sum of 2, = mean standard deviation of X variable scores 2, = mean standard deviation of Y variable scores N = number of measurements.

in which r

C

Figure 9 Negatively Skewed Curve

=

=

Product-moment correlation can help a researcher pinpoint the variance evident between two observations for each individual. In other words, the two variables are compared. When it is desirable to show the correlation between individuals by ranks rather than by variables, the Spearman rankdifference correlation technique may be used, applying this formula:

in which p = rank-difference correlation coefficient CD2 = sum of differences between ranking of the two variables squared N = number of measurements. Figure 10 is an example of rank-order correlation scores received by a group of high-school juniors on two music-aptitude tests administered by this ~ r i t e rRaw . ~ ~scores and ranks are listed for each test, and the differences between Ranks 1 and 2 are listed in D. Finally, D2 gives the Squares of the differences between each

Research in Music Education

188

Experimental Research

Figure 10 Rank-Order Correlation of Scores Obtained on Music Tests (Student

X,

X.

X,Rank

X,Rank

D

DZ

Another measure of variance is standard scores. When raw test scores are converted into standard scores, it is possible to "compare scores within a group arid between groups to add the scores from two or more tests to obtain a single score," according to T ~ c k m a nStandard .~~ scores are based on standard deviation units, or distance of the scores from the mean or X.The most useful measure of standard scores is the Z score. By definition, a Z score has a mean of 50 and a standard deviation of 10. The formula for Z scores is

Standard Scores

i

in which Z X

X *Subjects F and J both attained a score of 41 on X „ therefore share the mean of ranks 8 and 9. The Same is true for subjects B and E under X?.

pair. Based on data from Figure 10, the Spearman rank-order difference correlation for these scores is:

189

U

the standard score = student's raw score = mean of group = standard deviation of scores of group.

=

Inferential statistics deal with inferences made from samples Standard Error of the Meun to whole populations. Standard error of the mean is used to determine the accuracy of estimates of the population mean from the sample mean. Hardyck and Petrinovich state that "the standard error is the statistic we use to make the best estimate possible of the numerical values of the standard deviation of the random sampling distribution of means. The size of the sample directly affects the standard error; the larger the sample, the smaller the standard error of the mean. The formula for the standard error of the mean is

Experimental Research in which