The effect of information load and cognitive style on decision quality in a financial distress decision task

This study examines the impact of information load and individual differences in cognitive styles on auditors\u27 bankruptcy prediction accuracy. The effect of information load on auditor decision making is an important topic in the study of audit judgment because in many decision tasks, auditors are required to evaluate several pieces of information before reaching a decision. Prior research has shown that when decision makers attempt to integrate an excessive amount of information in their decision making process, the result may be sub-optimal decisions. This study draws upon the Schroder, Driver, and Streufert (1967) model that describes the information use patterns of individuals. Schroder and colleagues have shown that an individual\u27s information processing efficiency declines with excessive inputs of information and that information processing capability varies between individuals. This study also examines subjects\u27 confidence in their decisions. The study of decision makers\u27 level of confidence can be linked to the study of information load because it has generally been shown that increases in information lead to increases in confidence, while judgment accuracy correspondedly decreases with increases in information beyond some optimal point. In this study, forty-eight auditors made eighteen bankruptcy predictions at three different levels of information load. Each subject\u27s cognitive style was categorized as abstract or concrete as measured by the Kolb Learning Style Inventory. A 2 x 3 between-groups repeated measure design was used to examine the effect of information load and cognitive style on bankruptcy prediction accuracy. The results of the study were as follows: (1) a recency effect was found in some of the information processing, (2) the subjects had an overall low predictive accuracy, (3) seventy-eight percent of the subjects were classified as having an abstract cognitive style, (4) predictive accuracy did not follow an inverted-U pattern, (5) predictive accuracy differences between groups were not as expected, and (6) subjects\u27 confidence did not increase as the information load increased