観察学習におけるモデリング手がかりと代理強化および課題の複雑さ

大野木 裕明

doi:10.4992/jjpsy.49.137

MODELING CUES, VICARIOUS REINFORCEMENT, AND TASK COMPLEXITY IN OBSERVATIONAL LEARNING

HIROAKI OHNOGI

Author information

JOURNAL FREE ACCESS

1978 Volume 49 Issue 3 Pages 137-144

DOI https://doi.org/10.4992/jjpsy.49.137

Details

Published: September 20, 1978 Received: August 29, 1977 Available on J-STAGE: July 16, 2010 Accepted: - Advance online publication: - Revised: -
Correction information
Date of correction: July 16, 2010 Reason for correction: - Correction: ABSTRACT Details: Wrong : It has become evident through research on observational learning that attentional processes and vicarious reinforcement play important roles in what and how well an observer learns. But, many studies have failed to demonstrate the effects of vicarious reinforcement on observational learning. Bandura's position appears to be that a number of attention-controlling variables, some related to incentive conditions and others to the properties of the modeling cues themselves, will be influential in determining which modeling stimuli will be observed and which will be ignored. His theory has yet to be conclusively demonstrated.
The main purpose of the first experiment was to analyze the relationship between the effects of vicarious reward and the model's verbalization of the cues guiding the observer's future behavior. One hundred and sixty eight second grade children were divided into seven groups of subjects. A 3×2 factorial design containing levels of the model's verbalization (D₁, D₂, D₃) and vicarious reward (R, NR) was used (see Table 1). The purpose of the second experiment was to test the hypothesis that an observer learns not only specific cues from the model but a more generalized “rule”. One hundred and eight second grade children were divided into nine groups of subjects. A 2×2×2 factorial design was used containing task conditions of subjects (same, different), relevant dimensions of subjects (same, different) and vicarious reward (R, NR). But the number of dimensions (color, form) was held constant between the model's task and the observer's. Observers (subjects) performed a concept identification task after the observation of an expert model.
The main results were as follows: (1) The effect of observing an expert model was efficient (see Table 2). (2) Consistent with the hypothesis, comparison between R groups and NR groups in each D level showed that vicarious reward was not always necessary for facilitating the observer's subsequent imitative behavior, but the facilitation was possible in some conditions by heightening the observer's attention to modeling and associated situational cues (see Tables 2 & 3). (3) The main effect of the task condition was not significant, but the main effect of relevant dimension was significant (see Fig. 3, Table 4). The hypothesis that an observer learns a more generalized rule was almost supported. An analysis of response patterns was discussed with relation to multi-process in observational learning.
Right : Two experiments were conducted to examine the relationship between the effects of vicarious reward and a model's verbalization of the cues guiding an observer's future imitative behavior (in Exp. I) and the relationship between concept-identification task conditions for a model and an observer (in Exp. II). The main results were as follows: (a) Consistent with the hypothesis, vicarious reward was not always necessary for facilitating subsequent problem-solving, but the facilitation was possible in some conditions by heightening the observer's attention to modeling and associated situational cues. (b) The hypothesis that an observer acquires a more generalized “rule” was almost supported. An analysis of response patterns was discussed with relation to multiprocess in the observational learning theory.

Abstract

Two experiments were conducted to examine the relationship between the effects of vicarious reward and a model's verbalization of the cues guiding an observer's future imitative behavior (in Exp. I) and the relationship between concept-identification task conditions for a model and an observer (in Exp. II). The main results were as follows: (a) Consistent with the hypothesis, vicarious reward was not always necessary for facilitating subsequent problem-solving, but the facilitation was possible in some conditions by heightening the observer's attention to modeling and associated situational cues. (b) The hypothesis that an observer acquires a more generalized “rule” was almost supported. An analysis of response patterns was discussed with relation to multiprocess in the observational learning theory.

Corresponding author

Correction information

Register with J-STAGE for free!