Comprehension and Analysis of Information in Text. IV. Decision and Verification Processes.
Abstract
Reading texts for the purpose of making decisions was studied in a laboratory analog of a complex, natural, information-analytic domain. Subjects acting as stock brokers were trained in the first two sessions to categorize and evaluate stock report-type information. In addition, they learned to infer information from text-explicit facts. In Session three, subjects read texts and made decisions to Buy or Not Buy based on a conjuntive rule that was either given to the subject before (RB) or after (RA) reading the text. In Session 4, subjects read texts and then were presented probes that were to be verified as to either having or not having been presented in the previously read texts. Performance in all tasks was measured in terms of response latency, as well as accuracy. The results demonstrated that a specialized control schema for text comprehension develops that is based on the nature of the decision task. For example, readers in the RA versus the RB condition develop different text analysis techniques that are apparent in the decision, as well as the verification task performance. Most of the results are interpretable in terms of a model which suggests that purposes and goals relevant to the text comprehension process are incorporated into long-term memory (i.e., knowledge structures). The integration of these knowledge structures with current short-term information provides the basis of a task-appropriate text representation. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1981
- Accession Number
- ADA102677
Entities
People
- Ben Gurion
- Ely Kozminsky
- Lyle E. Bourne Jr.
- Steve Antos
Organizations
- University of Colorado Boulder