In this thesis, we develop and test theoretical models of the effect of two cognitive biases on performance appraisal (PA) accuracy. First, we build on the Constructivist Coding Theory to suggest that raters in low feedback availability environments mistake missing feedback for positive feedback, which causes them to develop overconfidence in their ability to provide accurate performance appraisals. We refer to this phenomenon as Raters’ Miscue (RM). Second, we integrate concepts from Social Exchange Theory (SET) and Feedback-Seeking Behavior (FSB) to argue that raters’ perception of direct feedback seeking leads to inflated performance ratings. We refer to this phenomenon as the Direct Feedback Seeking (DFS) effect. Design
We use a post-test only randomized experimental design to test our hypotheses. Over five experiments, we utilize a virtual laboratory environment to test the impact of errors in rater cognition on PA accuracy.Findings
First, we show that in low feedback availability environments, raters mistake missing feedback for positive feedback, which causes overconfidence in their ability to provide accurate appraisals (RM, Study 1). However, when we nudged raters to avoid misperception of missing feedback, the overconfidence effect attenuated (RM, Study 2). Second, our results suggest that raters perceive direct feedback seeking as a positive treatment, which creates an obligation for reciprocation with inflated ratings (DFS, Study 1). We show that performance appraisal purpose moderates the relation between feedback seeking and ratings (DFS, Study 2) and that Indirect Feedback Seeking reduces the effect of rater perception of feedback seeking on PA ratings (DFS, Study 3).Research implications
First, we contribute to a growing literature on the effect of cognitive biases on PA. Second, we study the development of overconfidence in PA, a topic that receives little scholarly attention. Third, we investigate the effect of feedback seeking on the rater, a perspective that has long been ignored in the literature.Practical implications
We make two recommendations to practitioners: 1) HR professionals must train raters to avoid misperception of missing feedback, and 2) multisource appraisal systems must be implemented with indirect feedback solicitation where the source of feedback seeking remains anonymous to the rater. Originality
To the best of our knowledge, this research is the first attempt to study the development of overconfidence in PA and to investigate the effect of perception of feedback seeking on PA ratings.Keywords
Performance appraisal; cognitive biases; overconfidence; constructivist coding; social exchange; feedback seeking; multisource appraisal systems.