Subculture, includes animation, comics, games (ACG), and idol fan culture, is popular among young generations in Japan. Previous studies have shown different psychological outcomes within different subcultural groups; however, underlying mechanisms remain unclear. This study proposes that subcultural identity may play a crucial role in mental health outcomes by interweaving social support and reputation. We examined the interplay between subcultural identity, social support, reputation, and different psychological outcomes through structural equation modeling (N = 300). Furthermore, we trained neural networks (NN) by applying a deep learning algorithm to predict psychological outcomes of different subcultures. The results suggest identity related to anime, idol, and hip-hop was positively associated with anxiety, aggression, depressive symptoms, and suicidal tendencies. By contrast, fashion and sports identities associated with no adverse or positive psychological outcomes. Perceived reputation mediates adverse psychological outcomes in the ACG, idol, and hip-hop groups. The highest accuracy in our NN reached 71%, indicating that NN could be an effective tool for predicting mental problems. Our work points up an urgent need to improve the mental health of the young generation by de-stigmatizing discriminated cultural groups.
Keywords: Neural networks; Psychological outcomes; Reputation; Social support; Subculture.
© 2022 Published by Elsevier Ltd.