Reliability and validity of a short Japanese version of the UPPS-P Impulsive Behavior Scale

This study aimed to verify the reliability and validity of a Japanese version of the S-UPPS-P Impulsive Behavior Scale. This is expected to facilitate comparisons of findings between international and Japanese samples in studies of impulsivity.

Two surveys were conducted. In the first survey, 632 participants, aged 20-44 years old, completed a translated version of the Japanese S-UPPS-P Impulsive Behavior Scale, the Motor Impulsiveness Scale, a short form of the Big-Five scale, the short Grit scale, and the brief version of the self-control scale. Two weeks later, the second survey containing the S-UPPS-P and the motor impulsiveness scale were completed by 450 participants who had completed the first survey to examine test-retest reliability.

In the first survey, an exploratory factor analysis was performed on the S-UPPS-P responses. A four-factor solution was the most suitable solution, with the factors of "Lack of Perseverance," "Lack of Premeditation," "Sensation Seeking," and "Negative-Positive Urgency." Then, a confirmatory factor analysis was performed. The conformity index of the original five-factor model was slightly better than that of the four-factor model. We also compared the five-factor model's conformity index with three other models that had been examined in the original and other foreign language versions of the S-UPPS-P. The five-interrelated factor model had the best model fit. The reliability of the five scales was confirmed. The scales exhibited internal consistency with α coefficients ranging from 0.65 to 0.79, in addition to the test-retest reliability ranging from 0.74 to 0.80. The convergent validity of each S-UPPS-P scale was supported by high relationships with the four personality scales, with the highest correlation coefficients ranging from 0.37 to -0.67.

The reliability and validity of the Japanese version of the S-UPPS-P were confirmed, despite the minor limitations of the exploratory factor analysis providing a four-factor solution instead of a five-factor solution, and the α reliability coefficients of two scales being acceptable but rather low. Thus, comparisons of findings between international and Japanese studies on impulsivity could be facilitated.