Multiple-Cue Probability Learningにおける幼児の手がかり使用について

土居 道栄

doi:10.4992/jjpsy.48.14

CUE UTILIZATION BEHAVIOR OF YOUNG CHILDREN IN MULTIPLE-CUE PROBABILITY LEARNING

MICHIE DOI

Author information

JOURNAL FREE ACCESS

1977 Volume 48 Issue 1 Pages 14-24

DOI https://doi.org/10.4992/jjpsy.48.14

Details

Published: May 20, 1977 Received: May 14, 1976 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 : The main purpose of the investigation was to study (a) the effect of cue dominance among multiple cues on the learning process of the multi-dimensional judgement tasks and (b) the facilitation of the multiple cue utilization in the task under the optimal condition of cue properties in six-year-old children.
Two experiments, based on the multiple-cue probability learning (MPL) paradigm, were conducted, and the data acquired were analyzed by means of the measures of lens model equation, achievement (γ_a), cue utilization (γ_s), knowledge (G), and cognitive control (R_s).
In these two experiments, the subjects were required to use two stimulus dimensions as cues to predict a lever distance (LD), a criterion which was determined by the formula LD=μk+νl, where μ and ν stand for the weight of cues of each dimension, and k and l for value assigned to the stimulus of each dimension respectively. The design was a 2×3 factorial with two cue dimensions, three levels of two cues validities (cue-criterion correlation). And the subjects were divided into six task conditions (see Tables 1 & 2) for each experiment.
In Exp. I, the effect of cue dominance in two cues on the performance related to the cue validity were examined, using the geometrical figures varied in size and brightness dimensions as the cue stimuli, assuming that the size dimension was the dominant cue.
In Exp. II, the children's ability to use multiple cue in MPL was examined under the condition that the factor of cue dominance was eliminated and that two available cues, size and length dimensions, were used for the task.
The major findings were as follows:
1) Measures of achievement and cue utilization revealed that the children had the tendency to use solely dominant cue in order to judge the multi-dimensional task. The performance of the task was a function of the dominant cue validity as the result of occurrence of uni-dimensional response to the dominant cue (see Figs. 2 & 3).
2) Under the condition of optimal cue properties, the six-year-old children were able to use multiple cues simultaneously in making judgement of MPL task corresponding to each cue validity (see Figs. 5 & 6).
3) The differences of the pattern of learning process between the measures of knowledge and cognitive control were observed which suggested that there were two different learning stages, that is, subject's detection of the task properties and their application of this knowledge to a given task, in the MPL process of children, as it was found with adults. The latter was more difficult to learn than the former for the children as compared with the adults' performances reported by the previous studies (see Figs. 4 & 7).
Right : The cue utilization behavior of six-year-old children in multi-dimensional judgement tasks was studied in two experiments based on the multiple-cue probability learning (MPL) paradigm. The task required subjects to use two stimulus dimensions as cues to predict a lever distance, a criterion which was determined by an experimenter. In Exp. I, the effect of cue dominance on the task performance were examined. The results showed that the children had a tendency to use solely dominant cue for the task, and the performance was a function of the cue validity (cue-criterion correlation) of dominant cue. In Exp. II, children's ability to use simultaneously multiple cues corresponding to the cue validity was confirmed, under the optimal condition of cue properties.

Abstract

The cue utilization behavior of six-year-old children in multi-dimensional judgement tasks was studied in two experiments based on the multiple-cue probability learning (MPL) paradigm. The task required subjects to use two stimulus dimensions as cues to predict a lever distance, a criterion which was determined by an experimenter. In Exp. I, the effect of cue dominance on the task performance were examined. The results showed that the children had a tendency to use solely dominant cue for the task, and the performance was a function of the cue validity (cue-criterion correlation) of dominant cue. In Exp. II, children's ability to use simultaneously multiple cues corresponding to the cue validity was confirmed, under the optimal condition of cue properties.

Corresponding author

Correction information

Register with J-STAGE for free!