Recently, studies in performance appraisal have focused more on the process of performance appraisal rather than on the mechanics of performance appraisal. This study focused on the individual differences among raters and the effect these differences have on recognition accuracy and rating accuracy. Specifically, individual differences were examined through individual constructs (personal prototypes) of different performance levels and individual levels of knowledge for a particular job. Furthermore, the effect of a competing task on recognition accuracy and rating accuracy were examined. The present study employed a 2 (presence/absence of competing task) X 2 (high family resemblance prototype/low family resemblance prototype) between subjects design with level of knowledge as a covariate. Eighty experienced carpentry workers and construction laborers from Southern Virginia were the subjects. The subjects were tested on their individual level of knowledge with a carpentry knowledge exam. Personal prototypes for each subject were obtained by individual interviews with each subject. The competing task manipulation consisted of subjects in the competing task condition receiving a task to complete while viewing the videotape while subjects in the no competing task condition did not receive a task. After the knowledge exam and prototype interview, the sample viewed a videotape of three carpenters performing four different woodworking tasks (sanding, sawing, hammering and staining) and completed a recognition and performance appraisal measure after observation. It was expected that subjects, regardless of their level of knowledge, with a competing task to perform during observation would make less accurate ratings than subjects with no competing task to perform. It was also hypothesized that raters with higher job knowledge would recognize more behaviors consistent with their personal prototypes and falsely recognize more behaviors from their personal prototypes than lower knowledge raters. Further, it was hypothesized that lower job knowledge raters should be more accurate in their recognition of ratee behaviors than would higher job knowledge raters. Further, it was hypothesized that higher knowledge raters would falsely recognize behaviors consistent with their personal prototype when there was a competing tasks than when was no competing task. There should be no difference for lower knowledge raters. Finally, it was hypothesized that high job knowledge raters would make less accurate performance appraisal ratings because of their reliance of their personal prototypes under increased processing conditions whereas there should be no such differences for lower job knowledge raters. The competing task did serve to divide the subjects attention in viewing the videotape as compared with subjects without a competing task. Little support was obtained for the hypotheses. These findings suggest that the presence/absence of competing task did not significantly effect rater accuracy and that subjects’ personal prototypes did not effect neither their recognition accuracy nor rating accuracy. Future implications for research study will be offered. / Ph. D.