Reducing Assimilation Effects in Performance Evaluation Judgments

Abstract

Recent research indicates that context effects play an important role in accounting decision making, where information for a given decision task is evaluated differently depending on the nature of the prior contextual information (e.g., Tan and Jamal 2001; Bhattacharjee, et al 2007). While these studies have contributed significantly to the literature, this research has focused on decisions where only prior task information is available to contextualize current task information. Yet, many accounting decision tasks also contain relevant benchmark information (e.g., budget targets) that can and should be used as the basis to evaluate current task data. We extend the research in this area by conducting a series of experiments in which the robustness of context effects from prior task information are tested in a setting where managers are supposed to compare current performance metrics against relevant and available target information. Consistent with our theoretical arguments, experiment 1 reveals that managers’ judgments using a basic scorecard are subject to context effects that result in assimilation such that an individual’s current performance is rated more favorably (unfavorably) when the individual’s prior performance was rated as favorable (unfavorable). This is the case even though in both conditions the same relevant benchmark information is provided and indicative of average performance for the individual. In experiment 2, we find that color status indicators, often used in practice and designed to make scorecard evaluations cognitively easier, are only partially successful at reducing assimilation effects in managers’ judgments. In experiment 3, we take an approach that is different from what has been used in prior scorecard research by having managers perform analytics on the data. This approach was inspired by the attention-directing aspect of analytics, popularized in the audit domain. This attention directing approach was found to successfully reduce assimilation effects. Our results provide insight into a different approach to scorecard judgments that may be beneficial for certain biases. Our results also extend the accounting literature on comparative-based judgments by demonstrating that prior contextual information can have such a strong influence on judgments that relevant comparative benchmarks (budgeted data) are ignored.