The purpose of this study was to replicate the Bentti, Golden, and Reigeluth study (1983), which explored the use of nonexamples to teach common errors as an effective strategy in teaching a procedure. A total of 24 undergraduate students enrolled in the Syracuse University Symphonic Band were randomly assigned to an experimental group and a control group. The experimental design was pretest-posttest, and two instructional booklets--one treble clef and one bass clef--were used to teach a procedure for building a musical interval up from a given note. The experimental group received instruction for the correct procedure plus reminders and examples of the most common mistakes; the control group received instruction for and examples of the correct procedure. Results showed no significant differences on an application level. The lack of significance may have been due to the fact that humans, as adaptive organisms, exert more or less effort to compensate for the quality of the instruction. Examples from the instruction booklet and posttest are included, and test scores are presented in tables and graphs. Sixteen references are provided as well as a list of prior publications in this series of working papers. (Author/MES)