Getting an Anchor on Availability in Causal Judgment

Abstract

The description of a problem in branching structures, such as fault trees and influence diagrams, can be helpful in estimating the contribution of possible causes to a problem. Previous research on the use of branching structures focused on the impact of omissions in the problem structure. Findings indicated that subjects tend to overestimate the importance of the branches they saw, and to underestimate the "catch all" category. The present study investigates the impact of replacing a factor by a number of subcategories belonging to this factor. Findings indicate that effects of these variations in the problem description are similar to the effects of omissions in the description, i.e., replacing a factor by a number of subcategories elevates the estimated contribution for the overall factor. The present study also investigates the effects on estimates of two related corrective techniques: generating additional (sub)categories and providing subjects with additional (sub)categories. The effects of the problem description were, however, not eliminated by providing subjects with additional information about the problem and by making the content of the factors more available by generating subcategories. Generating additional factors reduced the underestimation of the catch all category but this reduction was more pronounced if subjects also rated these self-generated factors. It is argued that availability defined as "ease of retrieval" is not sufficient to eliminate judgmental biases. Results indicate clear effects of availability based on "presence in the environment" and clear effects of anchoring and adjustment; each seems to explain part of the effects of the problem description on judgment.