The effects of interference, availability, and accounting information on investors' predictive judgments
UMI Dissertation Information Service eBooks, 1986
Prediction is one of the most important aspects of investment decision making. This study provide... more Prediction is one of the most important aspects of investment decision making. This study provides evidence that investors' predictive earnings judgments can be systematically influenced as a consequence of the combined effects of "output interference" and "Uavailability," and that the use of financial accounting information in the prediction process seems to provide limited benefit in terms of reducing this effect. Output interference is a psychological concept that implies that whatever is thought about first interferes with, and thus inhibits, later thoughts about an issue. An availability-based prediction strategy is one in which the decision maker uses the relative number of pro versus con reasons generated, and/or the ease with which such reasons can be generated, as cues in judging the likelihood of future events. Fifty-eight investors participated in an experiment that demonstrated that the order in which they considered opposing arguments regarding the possibility of reaching a specified level of earnings had an impact on both their ability to generate supporting and opposing reasons and their subsequent probability judgment that earnings would actually reach the specified level. The outcome for which the investors were able to generate the most supporting reasons was judged more probable. Investors were able to think of more reasons supporting a particular outcome, not because there were more such reasons in the objective environment, but rather as a consequence of output interference. The systematic effect on judgment, although perhaps slightly reduced, persisted when investors had access to financial statements while considering the company's earnings prospects. C ONSIDERABLE evidence presented in the judgment and decision-making literature indicates that individuals use heuristic procedures that simplify cognitive processing [Kahneman et al., 1982; Nisbett and Ross, 1980; Slovic et al., 1977; and Tversky and Kahneman, 1974]. In particular, people appear to employ cognitive heuristics when assessing or predicting the likelihood of uncertain events. Although these heuristic procedures are often efficient in that they reduce the complexities of such tasks, an extensive line of research indiI thank the members of my dissertation committee at the University of Wisconsin for their contributions to this research and for their professional guidance in general. This paper has also benefited from the helpful comments of my colleagues at the University of Pittsburgh, two helpful reviewers, and the participants of a University of Michigan accounting workshop. The research assistance of Sangho Do and Bonnie Morris, and the financial support of the Ernst & Whinney Foundation are gratefully acknowledged. Donald V. Moser is at The Joseph M. Katz Graduate School of Business, University of Pittsburgh. Manuscript received January 1988. Revisions received May 1988 and December 1988. Accepted December 1988.
Uploads
Papers by Donald Moser