Two studies examined whether the accessibility of performance prototypes influences performance appraisals. Pilot studies revealed students used performance prototypes when rating instructor performance. Study 1 manipulated the accessibility of these prototypes and the time delay of performance ratings. Results showed no effect of the prime on rating error and accuracy; however, discrimination accuracy decreased over time and recognition bias became more conservative. Study 2 manipulated prototype accessibility and type of rating stimuli (videotape vs. vignette). Rating accuracy and recall were higher for vignette than videotape stimuli, and only those participants exposed to the vignette exhibited priming effects. Results supported transfer-appropriate processing and implied that cognitive primes may have a stronger effect on performance ratings based on "paperpeople" than videotaped stimuli.The ability to effectively inhibit errors in performance appraisals requires an understanding of their root causes (Ilgen & Favero, 1985;Ilgen & Feldman, 1983). Consequently, research in the last decade has concentrated on cognitive processes by which raters collect, encode, store, and integrate ratee information in memory and retrieve it when needed (Cooper, 1981;DeNisi, Cafferty, & Meglino, 1984;Feldman, 1981). Obstacles encountered during any of these information-processing stages may undermine rating accuracy. The encoding of incoming stimuli about ratee behavior is one stage that is particularly vulnerable to rating errors (Nathan & Lord, 1983).During encoding, a rater compares perceived ratee features with schemas or prototypes (Fiske & Taylor, 1991;Rosch, 1975). When ratee features resemble a conceptual category, the ratee is assigned to this category (Feldman, 1981). In this form, a transformed version of original ratee information becomes amenable