Accuracy of Performance Measurement: An Investigation of Training Method and Amount of Practice

The purpose of this research was to investigate the effects of training method and amount of practice-and-feedback on the accuracy of performance ratings and behavioral observation. This research was a 3 x 3 factorial ANOVA design. Training method was comprised of frame-of-reference, cognitive modeling, and a no-training control group. Practice-and-feedback consisted of 0, 1, and 3 practice trials. Undergraduates (N = 99) were randomly assigned to one of nine experimental conditions. Each participant viewed and rated 7 videotaped interview simulations. The results for performance ratings indicated that (a) frame-of-reference training produced the most accurate ratings for elevation, differential elevation, and differential accuracy, and (b) practice-and-feedback did not improve accuracy. The results for behavioral observation revealed that (a) cognitive modeling training was effective in reducing the raters\u27 1-hit rates, (b) training method had no effect on false alarm rate, and (c) practice-and-feedback were ineffective for both observation error rates. Interpretation and suggestions for future research are discussed