OVERVIEW:
THE NATURE OF RESEARCH AND COMMON RESEARCH MODALITIES
IN MUSIC EDUCATION/MUSIC THERAPY
The scope or range of research in music education and music therapy is typically broader than that of musicology. Music education research, for example, focuses not only upon music, but also on education and the convergence of music and education. Likewise, research in music therapy is concerned not only with music, but particularly its use for non-musical purposes in various therapeutic contexts.
A musicologist, for instance, may investigate a Mozart symphony in a way that sheds light on its formal structure or the performance practices of its period. A music education researcher may investigate the same Mozart symphony with an eye toward determining, for instance, if listening to it today may impact children’s spatial reasoning abilities, or if certain ways of rehearsing and conducting the music result in more effective performance and musicianship. A music therapist, furthermore, might be interested in ascertaining under what conditions and with what clients Mozart's music may have a therapeutic calming or relaxing effect on patients undergoing chemotherapy.
In order to be “good research,” all three kinds of investigation must meet certain standards of rigor, objectivity and integrity. One is not better than the other; the three are simply different in terms of purpose and, often, methodology and reporting format.
Most current research in music education and music therapy reported in major U.S. journals and Dissertation Abstracts tends to be quantitative, both descriptive and experimental. Historical and philosophical research are growing. By comparison, qualitative research is still fairly new.
FIVE MAJOR RESEARCH MODALITIES IN MUSIC EDUCATION/MUSIC THERAPY
Historical Research
Historical research in music education/music therapy is a systematic inquiry that investigates, records, analyzes, and interprets events of the past for the purpose of discovering generalizations that are helpful in understanding the present and predicting the future. The focus of historical research is on change, growth, or development of individuals, groups, practices, movements, institutions, or ideas. Historical research in music education/music therapy attempts to go back to primary, original, or first-hand sources of information.
Rainbow & Froehlich (1987) point out that all historical research should follow four
methodological steps. “These steps are (a) to ask a multitude of precise questions related to particular events or person(s) of the past; (b) to gather and verify facts (data) related to those questions; (c) to develop criteria for selecting the data most useful in answering the questions and for rejecting information that may be considered irrelevant within the context of the study; and (d) to interpret the data by answering all questions posed” (p. 108).
Historical research describes what has happened in the past. Therefore, variables cannot be manipulated. True cause and effect cannot be determined, but historical research may sometimes be inferential.
According to Rainbow & Froehlich, there are five main categories of historical research as it relates to music education:
1.Chronology - Examine events and individuals on a time line; that is, to
observe phenomena as they occur and evolve over time.
Hypothetical Example: “Hand signs in the teaching of music from Guido to Curwen.”
2.Comparative - Comparison of events and individuals within specified time spans
but in differing spatial locations.
Hypothetical Example: “The Pestalozzian movement in Europe and in the United States, 1835-1845.”
3.Oral - Collecting data of an historical nature by means of interview. Procedures
include formal interviewing of a person, recording the interviews, interpreting
and preserving the transcribed interviews.
Hypothetical Example: “Wiley Housewright, Karl Glenn, and Will Schmidt: Interviews with three past presidents of the Music
Educators National Conference.”
4.Psychohistory - Includes procedures borrowed from the behavioral sciences;
uses both historical analyses and clinical insights.
Hypothetical Example: “John Philip Sousa, Manic-Depressive.”
5.Quantitative -Use of standard statistical data in support of historical research
questions.
Hypothetical Example: “Support of music organizations in Milwaukee, 1920-1950, as measured by donations, ticket sales, and concert attendance.”
Philosophical Research
Socrates’ practice of questioning “what everyone knows” (i.e., the status quo) remains a valid starting point for defining the role of philosophical research in music education. Philosophic research deals with questions involving what should be or what ought to be. In some sense, philosophic research encompasses quantitative, qualitative, and historical research. It is free to speculate (in a logical, systematic, and objective context, of course), thereby setting up hypothetical situations with their attendant hypothetical cause/effects, independent/dependent variables, inferences, interpretations, etc.
Schwadron (1973) neatly states the general purpose of philosophic research in music education: “Whether realistically or idealistically viewed, its operational sphere is in the realm of ideas; its motivation derives from an uneasiness with the status quo; and its functional end is that of preserving the integrity of music in education” (p 41).
Cady (1967) describes three basic procedures in philosophic research:
1.Analysis. To examine the meanings and relationships of phenomena. The analytic function searches out the implications of assertions, their consistency, and the assumptions involved in a body of theory.
2.Criticism. To examine basic alternative modes of life and thought. The formulation of these alternatives thus presents the basic choices that confront us and thereby the task of evaluating is created.
3.Speculation. To examine the imaginative and visionary aspect of philosophic studies in order to find constructions of ideal futures, projections or desirable societies, experiences, and ends for (hu)mankind (p 170).
Some examples of philosophic research in music education:
Daugherty, J.F. (1996). Why music matters: The cognitive personalism of Reimer and Elliott, Australian Journal of Music Education, 1(1), 29-37.
Reimer, B. (1995). Gender, feminism, and aesthetic education: Discourses of inclusion and empowerment. Philosophy of Music Education Review, 3(2), 107-124.
Quantitative Descriptive Research
Descriptive research is a research plan based on systematic observation of an existing situation. Variables are not manipulated by the researcher. Data are recorded, organized, and subjected to treatment in order to develop statistics and parameters that represent and describe the sample or population. For example, to determine that the average score of two hundred students taking an exam is 50 points (out of a possible 100 points) is a statistic that tells us something about the exam: It is too difficult for the students (unreliable or invalid).
Descriptive research cannot be inferential; and since variables cannot be manipulated, it is impossible to determine cause and effect. Sometimes, however, tests of significance are utilized to determine if real differences exist between two statistics or parameters. There are three basic categories of descriptive research designs, and each of these can be subdivided into more specific categories (Cady, 1967). Be advised, however, that usually a descriptive research project incorporates more than one category, or design.
1. Surveys: To collect detailed descriptions of existing phenomena with the intent of
employing the data to justify current conditions and practices, or to make more intelligent plans for improving social, economic, or educational conditions and processes. The objective may not merely be to ascertain status, but also to determine the adequacy of status by comparing it with selected or established standards, norms, or criteria.
a. Existing status: To collect detailed descriptions of existing phenomena
with the intent of ascertaining current conditions and practices or making judgments about social, economic, or educational conditions and
processes.
Hypothetical Example: “A survey of university music departments in the southeastern United States with respect to using the Kodaly method for teaching sightsinging.”
b. Comparison of status: To determine the adequacy of status by comparing
it with selected or established standards, norms, or criteria.
Hypothetical Example: “A survey of use of the National Standards for Music Education in Virginia high school music courses.”
c. Methods of improving status: To collect information from others as to how they have solved similar problems in order to obtain information that will assist in the improvement of an existing situation.
Hypothetical Example: “A survey of choral music education teachers whose programs have grown under block scheduling.”
2. Interrelational. To collect not only information about existing status, but also to
endeavor to trace interrelationships between the facts obtained to gain a deeper
insight into the phenomenon or phenomena.
a. Case study: To make an intensive investigation of the complex factors that
contribute to the individuality of a social unit --a person, family group,
social institution, or community.
Hypothetical Example: “The Moravians of North Carolina: A case study of music instruction in a religious community.”
b. Causal-Comparative: To discover not only what a phenomenon is like,
but if possible, how and why it occurs. Then compare the likenesses
and differences among phenomena to find out what factors or circum-
stances seem to accompany certain events, processes, or practices.
Hypothetical Example: “Comparison of SAT scores among students
enrolled in high school music performance classes and students not
enrolled in high school music performance classes.”
c. Correlation: To ascertain the extent to which two variables are related,
that is, the extent to which variations in one factor correspond with
variations in another. Typically involves multiple regression analysis.
Hypothetical Example: “Size of larynx and voice classification among adolescent males.”
3. Developmental: To be not only concerned with the existing status and interrelationships of phenomena, but also with changes that take place as a
function of time. Variables are described in the course of their development over
a period of months or years.
a. Growth: To determine the nature and rate of changes that take place in human
organisms.
Hypothetical Example: “Larynx size and voice classification among
females in childhood, puberty, adolescence and adulthood”
b. Trend: To obtain social, economic, or political data and analyze it to
identify trends and to predict what is likely to take place in the future.
Hypothetical Example: “Economic and political support for the National
Endowment for the Humanities, 1975-1995.”
Descriptive research utilizes non-parametric statistics (nominal and ordinal data), e.g., mode, median, mean, percentile, range, standard deviation, variance, z score, T score, Contingency Coefficient, Kendall Coefficient of Concordance, Spearman correlation coefficient, Chi square, Fisher test, regression equation, standard error of estimate, etc.
Descriptive research data are typically presented/reported in such formats as frequency distribution, contingency table, scattergram, histogram, frequency polygon, frequency curves, crosstabs, etc.
QuantitativeExperimental Research
Experimental research is a research plan undertaken to test cause and effect between independent and dependent variables. The independent variable is manipulated by the researcher to determine its effect upon the dependent variable. For example, once could test the effect of the independent variables of music and no music and the dependent variable of calculating math problems. Sixty music majors could be randomly assigned to the music and no music groups. The music could be a representative sample of contemporary orchestral music. A test of significance would be used to determine if the observed difference between groups with regard to solving the math problems correctly is real, or whether it occurred by chance. Then one would be able to state something to the effect that the probability that the difference in scores was caused by the independent variable (music or no music) is 95 to 5 (or: the difference is significant at the .05 level).
Experimental research usually is inferential, and usually examines only samples of the variable (since examination of entire populations usually is too massive a task).
Experimental research utilizes parametric statistics, (interval and ratio data), e.g., t-test, ANOVA, ANCOVA, multivariate analyses (MANCOVA, etc.), Pearson’s Product-Moment Correlation, etc.
Some Examples of Experimental Research:
Gregory, D. (1989). Using computers to measure continuous music responses. Psychomusicology, 8(2), 127-134.
Madsen, C.K., Standley, J.M., & Cassidy, J.W. (1988). Demonstration and recognition of high and low contrasts in teacher intensity. Journal of Research in Music
Education, 37, 85-92.
Standley, J.M., & Madsen, C.K. (1990). Comparison of infant preferences and responses of auditory stimuli: Music, mother, and other female voice. Journal of
Music Therapy, 26(4).
Miyasaki, K. (1989). Absolute pitch identification: Effects of timbre and pitch region. Music Perception, 7, 1-14.
Montgomery, A. P. (1996). Effects of tempo on music preferences in children in elementary and middle school. Journal of Research in Music Education, 44 (2), 134-146.
Pembroke, R. G. (1987). The effect of vocalization on melodic memory conservation. Journal of Research in Music Education, 35, 155-169.
Descriptive versus Experimental Research
Although descriptive research and experimental research are distinct and definable approaches to solving research problems, sometimes they are confused. The basic difference is one of cause and effect. A descriptive study looks for descriptions of existing phenomena, while an experimental study looks for the reasons why those phenomena exist in their present form (what causes the effect).
For example, a strictly descriptive study might be your evaluation of students in a sight singing class. You are interested in each student’s sight singing ability (the phenomenon), so you ask them, one by one, to sing one or more unfamiliar songs from notation. You score each person according to the number of pitches and rhythms he/she missed, and the result is a descriptive assessment (a score) of each student’s sight singing ability. You may want to go one step further and compare all class members statistically to see if their scores differ significantly from each other.
Notice that in the above example you were not looking for the cause of each student’s score. If this were your intention, however, you would want to “manipulate” the music. For example, you might believe that chromatic melodies are more difficult to sing than diatonic melodies--thus causing changes in the students’ scores. So you develop two sets of melodies that are equivalent in every way except that one set is diatonic and the other set is chromatic. You ask your students to sight sing both sets, and you score (missed pitches/rhythms) the two performances separately. Then you compare the two sets of scores for significant differences. In this case, then, you have manipulated the independent variable (diatonic/chromatic melodies) in order to pin down a cause for the effect of higher or lower scores in sight singing (the dependent variable). Probably you will find that chromatic melodies caused the students to earn lower scores than the diatonic melodies. Clearly, you have just completed an experimental study, because you investigated cause and effect).
Evaluation of Descriptive and Experimental Research
“Garbage in, garbage out.” The success of descriptive and experimental research studies depends on many factors, among them:
Research Questions:formulating logical, succinct, testable hypotheses or questions
The Research Design: arranging research materials and situations to minimize effects
of non-research variables. Poorly conceived research design
and lack of sufficient controls will yield meaningless data.
The purpose of a good research design is to allow for the
greatest control of the research situation by the researcher so
that he/she is confident that results are functions of the research
variables, not functions of other environmental conditions.
Statistical Design:statistical procedures applied to data collected from a particular
research design. The crucial and most difficult aspect of
statistical design is finding the one most appropriate to the
specific data and the purpose of the particular study.
Sample:A random sample is the most objective type of sample. In order
for all population elements to have an equal and fair (unbiased)
opportunity, they are randomly chosen (by specific and special
techniques) as elements to be included in the sample. If sample
elements have not been chosen in an unbiased manner, it is
dangerous to infer results to the entire population. A more
precise random sampling technique is to use a stratified sample
when the population is defined as consisting of two or more strata
(e.g., musicians, non musicians).
Good research is also well-reported research. Research articles in reputable journals contain the following elements that will help the reader: Introduction: (a) statement of the research problem; (b) review of related and background literature; (c) purposes of the present study; (d) specific research questions or null hypotheses formulated for the present study; (e) delimitations of the study, if any; (f) definition of terms, if any; (g) assumptions, if any. Method: (a) subjects; (b) apparatus/materials used; (c) procedures, including controls and research/statistical design. Results: clearly and appropriately stated for each research question or null hypthosis. Discussion/Summary : should relate to the results and practical value of the study, as well as suggestions for further research.
Qualititative Research
Various appellations are associated with qualitative research: case study, field study, ethnographic research, naturalistic research, phenomenological research, interpretive research, symbolic interactionist research, etc. According to Bresler and Stake (1992), qualitative research strategies “share certain characteristics: (1) noninterventionist observation in natural settings; (2) emphasis on interpretation of both emic issues (those of the participants) and etic issues (those of the writer); (3) highly contextual description of people and events; and (4) validation of information through triangulation” (the checking of data against multiple sources and methods) (p 76).
An early example of what today would be termed qualitative research in music education was a ten year research project funded by the Pillsbury Foundation (1941-1951) that sought to discover children’s musical development by observing and analyzing children’s free, unhampered musical play. Since that time, researchers have refined methodologies of qualitative research, using it particularly in complex situations where isolation of several variables is problematic. Bresler and Stake (1992) list the following weaknesses of qualitative research: (1) excessive subjectivity in observations; (2) imprecise language in descriptions; (3) vague descriptions of the research design; (4) unwieldy and voluminous reports; (5) implication of generalizability when little is warranted; (6) cost and time overrun; and (7) unethical intrusion into personal lives. Among the strengths of qualitative research are: (1) a holistic, systemic purview, emphasizing inner workings and contexts; (2) a strong, empirical commitment to triangulated description of teaching; (3) an obligation and opportunity to get the most from filed-work interpretations; and (4) a sense of empathy enhancing the utility of use for applied practice in education (p 87).