Abstract
This study aimed to develop a Japanese version of the Iowa Gambling Task (IGT) using Psychology Experiment Building Language based on the English version. Additionally, we tested the equivalence of these versions using a biological equivalence test. We randomly assigned 63 undergraduate and graduate students from several Japanese universities to our parallel design experiment. We evaluated equivalence by determining differences between the logs of the number of cards chosen from the advantageous decks in the Japanese and English versions in five blocks. We defined an equivalence margin as ± 30% (log (0.70 = -0.36, log (1/0.70) = 0.36, for log-transformation). All two-sided 90% family-wise confidence intervals in the five blocks were within the equivalence margin. Our results confirmed the equivalence of the Japanese and English versions of the IGT. The Japanese version can be used with the compiled executables and source code without charge. Future comparative studies should evaluate the IGT on Japanese samples or between samples from Japan and other countries.
