Examining Item-Content Relevance of the Dance Performance Evaluation Scale

Abstract

Background/Purpose: Examination of item content-relevance is the most important procedure in developing a measurement tool for evaluating a student's dance performance. Expert judges frequently rate the degree of match between the content of scale items and the objectives to be measured by the scale. Item content-relevance is estimated by calculating the percentage (i.e., 80%) of experts who classified the item into its proposed subscale. This approach provides information related to the content-relevance of individual items, but not the content representativeness of the set of content-relevant items. The purpose of this study was to examine item content-relevance of the Dance Performance Evaluation Scale (DPES) using the statistical procedure suggested by Aiken (1985) and to select the items which could evaluate the dance performance of various dance types in a valid way. Method: The 18-item DPES was administered to 20 Korean dance experts. The expert panel rated the degree of match between 18 items and five different dance types, such as Korean traditional dance, Korean creative dance, modern dance, traditional ballet, and creative ballet. Ratings were done using the following 5-point Likert scale: 1 (Poor Fit), 2 (Fair Fit), 3 (Good Fit), 4(Very Good Fit), and 5 (Excellent Fit). Mean and standard deviation, one-way MANOVA, and Aiken's V coefficient were calculated to examine item content-relevance ratings provided by Korean dance experts. Analysis/Results: The means of item content-relevance ratings among the five dance types were statistically significant difference from Item 1 (Originality of work, F(4, 16) = 9.43, p < .001), Item 9 (Proficiency of technique, F(4, 16) = 4.92, p = .01), Item 10 (Flexibility of body, F(4, 16) = 4.39, p = .03), and Item 18 (Appropriate physique as dancer, F(4, 16) = 3.07, p = .047). Similar results were identified between the means of item content-relevance ratings and the Aiken's V coefficients for the 18 items. With the exception of the rating for Item 1 of Korean traditional dance type (V = .56, p > .05) and traditional ballet type (V = .57, p > .05), the remaining 88 values of V were statistically significant at p < .01. Conclusions: The results showed that the student's performance for the different dance types could be evaluated by the items of DPES in a valid way.