American Accounting Association 2006 Midwest Region Meeting March 30 – April 1 Chicago, Illinois

Friday, March 31, 4:00 p.m. to 5:40 p.m.
Concurrent session 4A - Belief Revision and Evidence (Accounting, Behavior and Organizations)

Title: Examining Continuous Assurance Mitigation Effects on the Belief Revisions of Non-professional Investors: A Multi-Sample Approach

Robert Pinsker Old Dominion University

Bryan Church Georgia Tech University

ABSTRACT: Hogarth and Einhorn’s (HE; 1992) belief-adjustment (BA) model has been used as the theoretical framework to extensively study belief revision and order effects in both the psychology and accounting domains. In almost all cases, the model’s predictions have been supported. One limitation with this stream of research is that the focus has been on finding the recency order effect (i.e., more weight is placed on the most recent information relative to earlier information), rather than finding ways to mitigate it or better control the decision-making process that creates it.

Recency results from a significant amount of belief revision in the direction of the most recent information. If the magnitude of the belief revision could be reduced, so may the recency effect. According to the BA model, an individual’s sensitivity toward cues plays a large role in determining the magnitude of belief revision. Using source credibility theory, this paper attempts to use continuous assurance as a way to reduce an individual’s sensitivity to firm disclosures, which is then hypothesized to reduce the magnitude of belief revision.

The research methodology consists of a 2 (assurance: present or absent) by 2 (order: four positive disclosures followed by four negative disclosures or vice versa) by 3 (participant group: business professionals, MBA students, or undergraduate accounting students) between-subjects design. The results show that continuous assurance effectively mitigated belief revision magnitudes for all groups. Regardless of order, individuals receiving continuous assurance revised their beliefs significantly less than those not receiving assurance. If these individuals’ interactions in the marketplace result in increased stock price volatility, then using continuous assurance as belief revision mitigation techniques makes sense and has value. As such, the current paper adds to the existing continuous assurance and belief revision literatures.

Further analysis indicates undergraduate students had significantly greater belief revisions than the business professionals. This result can be explained as the novice investors (undergraduates) being relatively less familiar with the task and the role of assurance (compared to the business professionals) and, therefore, being relatively more sensitive to the firm disclosures. There were no differences between the MBA sample and the other groups. Given these results, researchers should exercise care when using a sample from one population (i.e., undergraduates or MBA students, business professionals) while attempting to explain non-professional investor behavior in the aggregate.

Back to the program