Reliability of scaling error detection

Reliability of scaling error detection among dental hygiene instructors is a basis for grade equity. This study tested statistically (ANOVA) grading reliability and variance factors using ten students' scaling performance scores while treating 34 patients. Nine instructors were grouped according to experience, and the patients were grouped according to four difficulty levels. The interinstructor differences were significant in the inexperienced group only. Variance analysis implied that 26 to 94 percent of the grade given is due to instructor differences rather than the students' performances. Experienced graduate students graded most consistently (26 percent of variance due to instructor differences) while faculty were less consistent (64 percent) and inexperienced graduate students exhibited marked inconsistency (94 percent). Error detection appeared to be more closely related to the instructors' educational background than to experience. Results of the study suggest the need for clinical rotation of instructors as well as for instructor calibration.