時間の長さの心理的分布について

神宮 英夫

doi:10.4992/jjpsy.48.337

THE DISTRIBUTION OF INTERNAL DURATION

HIDEO JINGU

Author information

JOURNAL FREE ACCESS

1978 Volume 48 Issue 6 Pages 337-343

DOI https://doi.org/10.4992/jjpsy.48.337

[Correction] The Japanese Journal of Psychology Vol.49 No.2 (1978) pp.110c-110c

Details

Published: March 20, 1978 Received: June 13, 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 : The purpose of this study was to obtain experimentally the distribution of internal duration of a fixed stimulus duration. Up to now, there have been two models for this distribution. They are Creelman's “neural firing counting model” (1962) and Allan et al.'s “onset-offset model” (1971). Creelman assumed the normal distribution, and Allan et al., the triangular distribution. Our problem was to find for which distribution the internal duration of 60msec would fit better.
Each of two subjects was trained to discriminate between two durations (30msec and 90msec). The distribution of reaction time (RT) was obtained from the two-alternative forced-choice judgment. It was assumed that these two alternatives were analogous to the two internal durations of 30msec and 90msec and formed a single criterion. By the application of the latency function hypothesis (Navon, 1975), 320 RTs obtained from each subject were divided into eight categories and this RT's distribution was converted into that of the internal duration. The coefficient of determination was used to evaluate the goodness of fit.
Two stimuli (a dot (condition NI), and a cross consisting of 17 dots (condition I)) were presented in random order, each for 60msec. The eight ranges of RT were regarded as identical for these two conditions in order to examine the difference of these conditions within each subject. This method using RT's distribution is distinctly different from such a method as duration discrimination used by Creelman and Allan et al.
The results were as follows:
1. Under the condition NI, the triangular distribution had better fit than the normal distribution.
2. Under the condition I, the triangular distribution also had better fit than the normal distribution. The goodness of fit, however, was worse than NI. In addition, the mean of distribution for I was less than that for NI, though NI and I both used the same duration of 60msec. These results suggest that the processing time of a cross pattern has an effect on the internal duration.
Right : There are two models for the distribution of internal duration; one assumes the normal, and the other, the triangular distribution. The purpose of this study was to examine for which distribution the internal duration of 60msec would fit better. In the forced-choice judgment of the duration of a dot or a cross, 320 reaction times (RTs) were obtained from each of two subjects. By the application of the latency function hypothesis (Navon, 1975), RT's distribution was converted into the distribution of internal duration. The results were: (a) The triangular distribution showed better fit than the normal distribution; (b) fit was worse when the stimulus was a cross rather than a single dot, which suggests that the processing time of cross (+) pattern has an effect on the internal duration.

Abstract

There are two models for the distribution of internal duration; one assumes the normal, and the other, the triangular distribution. The purpose of this study was to examine for which distribution the internal duration of 60msec would fit better. In the forced-choice judgment of the duration of a dot or a cross, 320 reaction times (RTs) were obtained from each of two subjects. By the application of the latency function hypothesis (Navon, 1975), RT's distribution was converted into the distribution of internal duration. The results were: (a) The triangular distribution showed better fit than the normal distribution; (b) fit was worse when the stimulus was a cross rather than a single dot, which suggests that the processing time of cross (+) pattern has an effect on the internal duration.

Corresponding author

Correction information

Register with J-STAGE for free!