The purpose of this study was to test the premise that feedabck intervention cues differentially direct attention to a level of a processing hierarcy proposed by Kluger and DeNisi (1996). The hierarchy consists of task details, task motivation (general task), or meta-task processes (attention to "self"). Feedback designed to initiate different level of processing was manipulated and performance on a typing task was measured. The relationship between the feedback manipulation and performance was analyzed through analysis of covariance and repeated measures analysis of variance. For the analysis of covariance, the assumption of equality of slopes was violated, so data were analyzed through an ATI design. The feedback manipulation was associated with changes in performance, and these changes depended on ability. / Master of Science