AUDITING

February 2000

WHEN ARE TWO AUDITORS BETTER THAN ONE? GROUP DECISION MAKING IN AUDITING

By Jean C. Bedard and James J. Maroney

In today's competitive audit environment, CPA firms must continually monitor their practices to ensure that their professional employees are used as efficiently and effectively as possible. Accordingly, the appropriate staffing of audit teams is an important consideration.

Many audit judgments are made by the team, as well as individuals. The amount of group interaction is normally left to the discretion of the audit team leader and is more likely to be a function of personal preference than professional guidance. In fact, there is little if any professional guidance in auditing that helps firm executives determine what types of audit decisions might be made most effectively at a group level while providing recommendations on improving the group decision process.

When should group decision making be used?

Psychology research has found that group decision making generally results in performance gain over individual decision making. However, since audit firms face a highly competitive environment, further specific guidance is necessary to determine which audit tasks justify the additional time required by a group decision process.

While relatively few studies address auditing specifically, two recent ones provide some assistance in this area. A 1998 study found that groups of auditors were much better than individuals at identifying potential financial statement errors but were no better at evaluating those errors. Another study found that groups identified more issues relevant to the disclosure decision than individual auditors. However, there was no difference in the final decision outcome.

A common finding from both studies was that the primary benefit from group decision processes was in the generation of relevant information, not in the evaluation of that information. Furthermore, a 1972 model of interactive group decision making concludes that the failure of group decision making is not usually due to inability to identify a good solution, but rather it is due to failure to select that "good" solution.

Audit firm executives can benefit by using this research to identify those audit tasks that are most appropriate for group decision processes. Since the primary benefit is information generation, complex audit tasks that require the consideration of a number of alternative ideas appear to be the types of tasks for which group decision processes might be most effective. For example, if unexpected fluctuations in financial statement balances are caused by a financial statement error, the failure to consider that explanation during analytical procedures can significantly reduce audit effectiveness. Thus, group meetings can be an effective and efficient way to generate potential explanations for unexpected fluctuations identified through audit analytical procedures.

Explanations can be generated at a group level and then evaluated at an individual level. While auditors often initially seek explanations from management for unusual fluctuations, a November 1997 Journal of Accountancy article discussed some of the pitfalls of this practice and suggests that auditors first develop independent explanations.

Research suggests that the following audit tasks would also be conducive to group decision processes:

* Identification of inherent risk, fraud risk, and control risk factors during audit planning.
* Identification of going concern issues and potential mitigating factors.
* Identification of issues relevant to selection of the appropriate form of audit opinion.
* Identification of issues relevant to footnote disclosures in the financial statements.
* Management advisory or tax projects that require the generation of a number of alternative ideas.

All of the aforementioned tasks share one thing in common: For these tasks to be performed as effectively as possible, all relevant factors must be considered. Furthermore, for several of these tasks, failure to consider a relevant factor can impact other phases of the audit. For example, failure to consider all relevant risk factors during audit planning would most likely result in an ineffective audit program and possibly in audit failure. Thus, the extra time required for group processing during the risk identification phase of audit planning may be well worth the investment.

How can group leaders enhance effectiveness?

To enhance the group decision process, group leaders must understand the potential advantages and disadvantages of group decision making. An extensive body of research in psychology has focused on the benefits (process gains) and costs (process losses) of group decision making. While relatively sparse, a few auditing studies have also begun to focus on the benefits and costs of the group decision process. Examples of some of the more important benefits and costs (in addition to the extra time required) are summarized in the Table .

While group decision making often provides process gain, psychology research has generally found that interactive groups rarely meet their full potential due to process loss. The 1998 study previously discussed found that interactive groups of three audit seniors were much better than individual auditors at identifying potential financial statement errors during analytical procedures. However, they were no better than individual auditors at evaluating these errors, primarily due to process loss. Specifically, interruption of one group member by another member prevented several groups from selecting a good solution. The groups that performed best were the groups that had few interruptions (members were allowed to explain the rationales behind their ideas) and were able to correct each other's reasoning errors.

Audit firm executives should realize the risks involved in group decision making and take steps to reduce the likelihood of process loss. While this may be easier said than done, the following suggestions may help:

* Group leaders should guard against domination of the discussion by one individual.
* Group leaders should create an environment where members are free to express and explain their ideas without interruption. However, if a group member has made a reasoning or factual error during the discussion, another group member should be allowed to interrupt.
* Group members should be encouraged to write down any ideas that come to mind while another member is speaking, so that they may bring them up later.
* A group member should record all ideas considered during the discussion. It is possible that further analysis may reveal that some little-discussed ideas deserve more attention.

Does group composition affect the quality of interaction?

Numerous psychology and sociology studies have examined the effect of gender on group interaction. Many of these studies have been based on "expectations status theory." Basically, this sociological theory predicts that gender and rank are considered to be status characteristics that can lead to different expectations as to how well certain individuals will perform a decision-making task. These expectations may lead to interaction and performance differences during group interaction. Since status characteristics are created and perpetuated through societal influences, most individuals are unaware of them.

Findings from the gender studies have generally been consistent with expectations status theory and have found that men tend to participate more than women in group discussions and, furthermore, men tend to interrupt women more than women interrupt men. While such research in auditing has been limited, a recent study used professional auditors as subjects and examined the effects of gender on group interaction during an analytical procedures task. The study found that women that worked in predominantly male group tended to verbalize less than women that worked in predominantly female groups and less than men regardless of group composition. As a result, the women in predominantly male groups contributed less than they could have to their group's decision process.

Since most audit teams consist of both genders, findings from this line of research have important implications for audit practice. Furthermore, since most audit teams are hierarchical (e.g., manager, senior, and staff) and rank is also considered a status characteristic, individuals at lower levels within an audit team may also interact less during group discussions.

As mentioned earlier, the effect of status characteristics on group interaction does not arise from conscious actions--most individuals are unaware that such effects exist. Therefore, if group leaders are aware that gender and rank within the firm can affect a staff member's interactions during group discussions, they can take steps to mitigate potential detrimental effects. Group leaders should--

* create an open environment where individuals can express their ideas,
* allow individuals an opportunity to defend their ideas, and
* be cognizant that certain individual characteristics may affect group interaction. Extra efforts may sometimes be required to encourage female and lower ranking group members to contribute their ideas to the discussion. *