Information overload research in accounting: a systematic review of the literature

Abstract

This paper reviews literature in the domain of information overload in accounting. The underlying psychological concepts of information load (as applied in accounting research) are summarized, and a framework to discuss findings in a structured way is proposed. This framework serves to make causes, consequences, and countermeasures transparent. Variables are further clustered into major categories from information processing research: input, process, and output. The main variables investigated are the characteristics of the information set, especially the number of information cues as an input variable; the experience of the decision-maker, the decision time, decision rule, and cue usage as process variables; and measures related to decision quality (i.e., accuracy, consensus, consistency) and related to selfinsight (calibration, confidence, feeling of overload) as output variables. The contexts of the respective research papers are described, and the operationalization of variables detailed and compared. We employ the method of stylized facts to evaluate the strength of the links between variables (number of links, direction and significance of relationship). The findings can be summarized as follows: most articles focus on individual decision-making in the domain of external accounting, with financial distress predictions constituting a large part of these. Most papers focus on input and output variables with the underlying information processing receiving less attention. The effects observed are dependent on the type of information input and the task employed. Decision accuracy is likely to decrease once information load passes a certain threshold, while decision time and a feeling of overload increase with increasing information load. While experience increases decision accuracy, the results on decision time and consensus are conflicting. Most articles have not established a significant link between changes in information load and changes in decision confidence. Relative cue usage, consensus, consistency, and calibration decline with increasing information load. Available time has a rather positive effect on decision accuracy and consensus. Based on these findings, implications for practice and future research are derived.