パターンの良さ判断とパターン内変換構造 パターン認知に関する今井の変換構造説の検討

松田 隆夫

doi:10.4992/jjpsy.49.207

JUDGMENTS OF PATTERN GOODNESS AND INTRA-CONFIGURATIONAL TRANSFORMATION STRUCTURES

AN EXAMINATION OF IMAI'S THEORY

TAKAO MATSUDA

Author information

JOURNAL FREE ACCESS

1978 Volume 49 Issue 4 Pages 207-214

DOI https://doi.org/10.4992/jjpsy.49.207

Details

Published: November 20, 1978 Received: June 07, 1978 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 : An experiment on judgments of pattern goodness was conducted to examine Imai's theory of pattern cognition which predicts the goodness changing systematically depending upon the intra-configurational transformation structures, and also to demonstrate some difficulties to be resolved.
Forty-eight prototypical patterns, each consisting of five solid circlets located in a 3×3 matrix of which no rows or columns were left vacant, were selected out of 90 such patterns that are possible, and they composed stimulus patterns for Conditions I to IV; which were specified by the presence or absence of four open circlets in complementary cells, and by the presence or absence of lines marking off cells (Fig. 1). In all, 192 stimuli were prepared. The structures of patterns were defined by invariant properties for such intra-configurational transformations as the mirror-image transformations on either left or right diagonal axis, M_D, the horizontal axis, M_H, or the vertical axis, M_V, and the 180°-rotation transformation, R. Accoding to these structures, 48 patterns were classified into six types: Two patterns were invariant for all the transformations (OR-joint transformation), 20 for only M_D, 4 for M_H, 4 for M_V, 4 for R, and the remaining 14 for none of them (Empty-conjugate, E), in which M_D, M_H, M_V, and R all refer to unitary transformations.
Thirty undergraduates served as subjects, and were required to make goodness judgments for 192 stimulus patterns on a 7-point scale, with 7 being the best and 1 being the poorest.
Table 1 shows mean goodness ratings for all stimuli. Main results were as follows:
(1) Goodness ratings for 48 patterns were significantly correlated among Conditions I to IV (Table 2), and the correlations obtained from the individual data were also high. Averaged ratings in I to IV were nearly the same.
(2) Pattern goodness was ranked in the order predicted by the theory; that is, patterns bearing the OR-joint transformation were rated highest, patterns with the unitary transformation were next highest, and patterns with no transformation structure were poorest. The differences among them were significant.
(3) There were also significant differences among mean goodness ratings of M_D-, M_H-, M_V-, and R-patterns excepting those between M_H and R, all belonging to the unitary transformation structure. Their descending order of goodness was M_V, M_D, R, and M_D.
(4) Among individual patterns in a given type of transformation structure, there was no significant difference, except that ratings of M_D-patterns were widely ranged from those higher than R-patterns to those nearly equal to E-patterns. The differences in M_D, however, were only recognized between pairs of patterns which had different inter-configurational transformation structures.
(5) Some subjects rated the goodness in the reverse order to the general tendency predicted. In a sense, this seems to be another evidence showing that the intra-configurational transformation structures play an important role in the cognitive judgment of pattern goodness.
Considering the results mentioned above, it seems that the present study supported Imai's theory to some extent. However, some difficulties remain and wait for future clarification; that is, the theory must be refined so as to explain the differences observed among types in the unitary structure and among patterns in the same type, and also the individual differences such as the transformation structures functioning in the opposite direction in performing the cognitive judgments of patterns.
Right : To examine Imai's theory which predicts the pattern goodness changing systematically depending upon intra-configurational transformation structures, 30 subjects were required to make goodness judgments for 192 stimuli, each consisting of five solid circlets in a 3×3 matrix with or without four open circlets in complementary cells and/or lines of demarcation. Mean goodness ratings supported the theory to some extent, but a few fundamental difficulties remained; that is, the theory must be refined to explain the differences observed among different types in the unitary transformation structure and among patterns in the same transformation type, and also the individual differences such as the transformation structures functioning in the opposite direction in performing the cognitive judgment of patterns.

Abstract

To examine Imai's theory which predicts the pattern goodness changing systematically depending upon intra-configurational transformation structures, 30 subjects were required to make goodness judgments for 192 stimuli, each consisting of five solid circlets in a 3×3 matrix with or without four open circlets in complementary cells and/or lines of demarcation. Mean goodness ratings supported the theory to some extent, but a few fundamental difficulties remained; that is, the theory must be refined to explain the differences observed among different types in the unitary transformation structure and among patterns in the same transformation type, and also the individual differences such as the transformation structures functioning in the opposite direction in performing the cognitive judgment of patterns.

Corresponding author

Correction information

Register with J-STAGE for free!