A Multivariate Analysis of Nonperformance Variables as Predictors of Marching Band Contest Results (Dissertation)
The purpose of this study was to examine a set of nonperformance variables pertaining to high school marching bands, and the relationships between these variables and the scoring results of marching band contests in the United States. The term "nonperformance variable" refers to a characteristic of a competing marching band or its director that was not overtly evaluated by performance criteria at a contest. The need for this study results from the widespread and historically established nature of band competitions, and evidence in prior research demonstrating pressure to compete. This study merged single-state methods and limited sets of variables employed in previous research by using a multi-state sample to examine a set of 22 nonperformance variables in a multivariate design.
To create a generalizable sample that could accurately reflect high school marching band activity around the United States, a sample of six states was randomly selected from a stratified design to avoid bias for geographic region and population size. Using an online questionnaire, data were collected from marching band directors whose bands participated in representative contests in the selected states, resulting in a response rate of 52.15%. Results of a stepwise multiple regression analysis indicated five significant models predicting 50.0% of the variance in contest scores. The five significant predictor variables included the size of the marching band; the number of uncertified paid assistant instructors; a scale of the director's attitudes towards marching band and competition; the hours of weekly rehearsal; and, the dollar budget of the marching band. All five predictors were positive, and all exhibited log-linear or inverse-linear relationships to the contest score, indicating a complex relationship.
The high percentage of variance in contest scores explained by this set of nonperformance predictors (50.0%) carries strong implications for participants and administrators of band contests, and raises issues of equity and fairness in the competitive process, particularly when considering that these variables are often largely beyond the control of the student performers. Further attention to these issues is suggested in research and practice, with a goal of ensuring that competitive outcomes are maximally determined by actual student performance.
David A. Rickels. "A Multivariate Analysis of Nonperformance Variables as Predictors of Marching Band Contest Results (Dissertation)" 2009
Available at: http://works.bepress.com/david_rickels/3