Uncertainty Measurement in a Complex Task as a Function of Response Mode and Event Type Characteristics.
Abstract
This research was concerned with the hypothesized effect of response mode and event type factors on measured uncertainty for frequentistic events observed within a realistic task setting. Subjects served as dispatchers of emergency services for a hypothetical city on five daily shifts. Events were generated by a stationary stochastic process (emergency calls) or by the subject's responses to these calls (allocations and correct responses). At the conclusion of the 600 dispatching trials, subjects in the two experimental groups estimated either the frequency (FE) or probability (PE) of various kinds of events, and in a subsequent session, made predictive choices among selected event pairs. A control group (C) followed the same procedure except for omission of the estimation task. Results showed that estimation performance is influenced reliably by both variables: the FE group was superior to the PE group on all events, externally generated events produced generally better estimates than internal ones, and spatial events were judged more accurately than non-spatial ones. Experimental groups made better choices than the control group, and there was a tendency for choice performance to reflect the quality of estimation performance. Possible explanations for the influence of estimation on choice measures are discussed together with various practical, theoretical, and research implications. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1981
- Accession Number
- ADA096687
Entities
People
- Shanta P. Kerkar
- William C. Howell
Organizations
- Rice University