Task Characteristics in the Formation and Use of Uncertainty Impressions.
Abstract
Fourteen studies ranging from narrowly focused molecular experiments to more broadly focused molar ones were completed on issues surrounding the general question of how people form and use impressions of event uncertainty. Particular attention was directed toward the influence of task features distinguished through an earlier taxonomic analysis. Our general hypothesis was that response demands, the defining characteristics of observed events, and prior beliefs regarding event causation play an important role in (a) how frequentistic evidence is processed, and (b) how accurately subsequent judgments or choices reflect the observed evidence. The overall findings were consistent with this hypothesis and therefore supportive of the taxonomy, although a number of questions remain concerning the underlying cognitive processes. For example, probability estimation is consistently less accurate than frequency estimation for the same observed events, and either type of estimation enhances the quality of subsequent choice behavior. We were able to show that the inferiority of probability to frequency estimation is not attributable to differential encoding and probably involves reliance on different kinds of stored information. However, we were not able to pinpoint what that information is. Similarly, we showed that the enhancement of choice by prior estimation is a general cuing phenomenon, but were not able to describe precisely how such cuing works.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1981
- Accession Number
- ADA108047
Entities
People
- William C. Howell
Organizations
- Rice University