Journal article
Tests of an all-or-none model of verbal mediated responding
Journal of experimental psychology, Vol.96(2), pp.247-254
12/1972
DOI: 10.1037/h0033661
Abstract
Investigated choice behavior in a multiple-choice task requiring S to choose the alternative that "went with" a stimulus word, where 1 alternative may be chosen on the basis of an experimentally induced mediator. A mathematical model is proposed as an initial attempt to understand how the mediation process proceeds over trials. The main assumption of the model is that Ss come to use the same mediator consistently from 1 trial to the next in an all-or-none manner. Exp. I with 46 undergraduates used a double-function list of common nouns for the acquisition stage and included a mixed- and an unmixed-list presentation of test items. Exp. II with 43 Ss used a 3-stage chaining paradigm with stages of the type paralog-noun, noun-trigram, and paralog-trigram, in which the 3rd stage (test list) contained either 2 or 3 response alternatives. Reliable mediation effects were obtained in each experiment, and the all-or-none assumption was supported in each experiment. The number of free parameters necessary to account for the data was determined for each experiment. The effects of variables of interest are related to specific parameters.
Details
- Title: Subtitle
- Tests of an all-or-none model of verbal mediated responding
- Creators
- Kent L Norman - U. IowaIrwin P Levin
- Resource Type
- Journal article
- Publication Details
- Journal of experimental psychology, Vol.96(2), pp.247-254
- DOI
- 10.1037/h0033661
- ISSN
- 0022-1015
- eISSN
- 1946-1941
- Publisher
- American Psychological Association
- Number of pages
- 8
- Language
- English
- Date published
- 12/1972
- Academic Unit
- Marketing; Psychological and Brain Sciences
- Record Identifier
- 9984963207802771
Metrics
1 Record Views