TRAINING EFFECTS ON ACCURACY OF PERFORMANCE APPRAISAL ACROSS METHODS, CONTENTS, AND FORMATS

This dissertation reviewed recent research involving training methods, training contents, and performance appraisal formats used in attempts to attenuate psychometric errors, along with research involving the same independent variables and their effects on accuracy in performance appraisal. Following a series of criticisms which focused on the use of small groups of untrained raters to determine true scores , results were presented from an experiment which contrasted two training methods (lecture versus group discussion), two training contents (accuracy versus psychometric errors), three performance appraisal formats (a behavioral expectation scale, a graphic rating scale, and a modified behavioral observation scale), and their effects on accuracy across three levels of performance (low, average, and high). A unique operational definition of accuracy was used, with results interpreted in light of a four-way interaction