THE DEVELOPMENT AND VALIDATION OF AN INSTRUMENT TO MEASURE WIND ENSEMBLE ERROR DETECTION SKILLS AMONG INSTRUMENTAL MUSIC EDUCATORS

Abstract

The Instrumental Music Error Detection Test, or IMEDT, was developed to examine error detection ability regarding pitch rhythm and articulation errors in recordings of a wind ensemble. This test was designed to simulate an authentic rehearsal situation. The musical excerpts were selected from grade three band literature and performed with full instrumentation. A total of 30 errors was inserted into the recordings; 12 pitch errors, 12 rhythm errors, and 8 articulation errors. A university wind ensemble recorded the excerpts, first as written, or what was considered to be a "model performance," and a second time with the errors inserted. The completed IMEDT contained two recordings of each of the eight musical excerpts, the first as written and the second with inserted errors.

The IMEDT was administered in six different test administration variations to determine the method that was most valid and reliable and had the highest internal consistency. Each test was administered in an individual setting with the participant and me and took approximately 45 minutes to an hour to complete. Sixty two participants completed this first phase of test administration. Using Cronbach's alpha to estimate the reliability and internal consistency, it was empirically decided that the test administration variation of score and recording with non-controlled time (S&R/N) had the highest alpha level. The order of musical excerpts was also determined empirically through this statistical test. Twenty additional participants completed the second phase of test administration of the IMEDT in the S&R/N method, again in a individual setting, taking approximately 40-45 minutes to complete. After data collection was complete, it was determined that the IMEDT was both reliable and internally consistent (á = .72).