[Prior research indicates that analysts' forecasts of earnings tend to be optimistic. Analysts' optimism may be attributed to experience, cognitive information search strategies, motivational incentives or some combination thereof. In this study, we conduct an experiment that uses a computerized eye-movement retinal imaging system to capture the cognitive search strategy of 60 professional financial analysts. We find that, within the experiment, more accurate analysts employ a directive information search strategy, whereas less accurate analysts employ a sequential search strategy. Experimental results also indicate that motivational incentives intensify the analysts' tendency to provide optimistic earnings forecasts. We also conduct an examination of the analysts' predictive accuracy outside the experimental setting. We find a significant relation between historical accuracy and the analysts' cognitive search strategy observed in the experiment. Post-experiment survey results provide insight into the linkage between specific accounting information used by the analysts and the accuracy of their forecasts.]

Abstract Prior research indicates that analysts' forecasts of earnings tend to be optimistic. Analysts' optimism may be attributed to experience, cognitive information search strategies, motivational incentives or some combination thereof. In this study, we conduct an experiment that uses a computerized eye-movement retinal imaging system to capture the cognitive search strategy of 60 professional financial analysts. We find that, within the experiment, more accurate analysts employ a directive information search strategy, whereas less accurate analysts employ a sequential search strategy. Experimental results also indicate that motivational incentives intensify the analysts' tendency to provide optimistic earnings forecasts. We also conduct an examination of the analysts' predictive accuracy outside the experimental setting. We find a significant relation between historical accuracy and the analysts' cognitive search strategy observed in the experiment. Post-experiment survey results provide insight into the linkage between specific accounting information used by the analysts and the accuracy of their forecasts.

This study complements and extends prior decision aid (DA) research by examining the DA reliance behavior of professional buy-side financial analysts in the context of their actual work environment. A large mutual fund company provided data on buy-side analysts' earnings forecasts over the course of one year, during which forecasts were made at the end of each quarter for the following four consecutive quarters. As part of the decision process, all analysts could voluntarily access a DA to assist them in forecasting earnings. Consistent with extant DA theory, the results indicate that analysts with greater performance-contingent incentives were less likely to rely on the DA and analysts with more complex portfolios were more likely to rely on the DA. Contrary to the results of most DA research and inconsistent with extant DA theory, analysts with greater task ability relied more on the DA than analysts with lesser ability. Finally, when DA reliance was high, analysts' forecast accuracy was also high, regardless of DA accuracy. The results provide valuable insight into the use of DAs by professional decision makers and the influence of DA reliance on their judgments in light of real-world pressures and performance consequences. The theoretical and practical implications of this study call for more research into why, how, and under what conditions highly skilled knowledge workers rely on the advice of DAs.

ABSTRACT: The current study examines the outcomes of three fraud brainstorming procedures—nominal group, round robin, and open discussion—via a randomized between-participant field experiment involving 150 audit clients and 2,614 auditors who participated in natural, hierarchical audit teams. The results indicate that nominal group and round robin brainstorming resulted in equivalent numbers of unique fraud risks and comparable increases in planned audit hours, while open discussion brainstorming yielded the least number of unique ideas and the smallest increase in planned audit hours. Furthermore, nominal group and round robin brainstorming yielded more changes/additions to the nature and timing of substantive testing than open discussion brainstorming. Study findings offer theoretical and practical insight into fraud brainstorming.