Scholarlink

<p>In an era of rapid change, the immense time pressure on graduate students has compelled them to turn to artificial intelligence (AI) tools, which have the potential to foster addiction. This study examines how perceived time pressure contributes to AI addiction among graduate students through academic control deprivation and how self-reflexivity moderates this pathway. A cross-sectional survey of 1257 Chinese graduate students was analyzed using latent-profile analysis and multi-group structural-equation modeling. Latent profile analysis (LPA) identified three student learning engagement profiles (low, moderate, and high engagement). Data were analyzed using structural equation modeling (SEM) with bootstrapping to test the moderated mediation model. Results show that perceived time pressure significantly increases AI addiction, partly via academic control deprivation, while self-reflexivity mitigates this indirect effect. The moderated mediation was strongest among high-engagement students, suggesting a paradox of vulnerability. These findings extend the self- vs. external regulatory theory by demonstrating how the interaction between contextual stress and self-regulation shapes AI addiction, offering insights for targeted interventions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

How time pressure intensifies artificial intelligence addiction among graduate students: exploring the role of academic control deprivation and self-reflexivity across engagement profiles

  • 期刊论文

  • 发表日期：

    2025年10月13日

查看全文
  • Yan Li,
  • Jia Jiang

摘要

In an era of rapid change, the immense time pressure on graduate students has compelled them to turn to artificial intelligence (AI) tools, which have the potential to foster addiction. This study examines how perceived time pressure contributes to AI addiction among graduate students through academic control deprivation and how self-reflexivity moderates this pathway. A cross-sectional survey of 1257 Chinese graduate students was analyzed using latent-profile analysis and multi-group structural-equation modeling. Latent profile analysis (LPA) identified three student learning engagement profiles (low, moderate, and high engagement). Data were analyzed using structural equation modeling (SEM) with bootstrapping to test the moderated mediation model. Results show that perceived time pressure significantly increases AI addiction, partly via academic control deprivation, while self-reflexivity mitigates this indirect effect. The moderated mediation was strongest among high-engagement students, suggesting a paradox of vulnerability. These findings extend the self- vs. external regulatory theory by demonstrating how the interaction between contextual stress and self-regulation shapes AI addiction, offering insights for targeted interventions.