Scholarlink

<p>The purpose of this study was to test whether student resilience moderates the relationship between perceived parental authority (authoritative, authoritarian, permissive) and academic performance in pervasive, AI-supported secondary-school learning environments, and whether human–AI interaction behaviours (shared metacognition, cognitive offloading) mediate these effects. The design of this study was a cross-sectional survey of 447 Grade 10–12 students (Guangdong, March 2025) in schools using an always-available, classroom AI tutor. Parenting styles (PAQ), resilience (school resilience scale), and AI-use constructs were measured; academic performance was self-reported (0–100) with robustness checks using quartile rank. We estimated ordinary least square (OLS) models with resilience interactions, a structural equation model (SEM) for mediation via shared metacognition and cognitive offloading, school-clustered SEs, and extensive sensitivity analyses. Authoritative parenting predicted higher grades (≈ + 2.1 points, <i>p</i> &lt; 0.01); authoritarian predicted lower (≈ − 1.5, <i>p</i> &lt; 0.05); permissive was null. Resilience strengthened the authoritative effect and buffered the authoritarian penalty (interaction terms ≈ + 1.0, <i>p</i> &lt; 0.05). Resilience increased AI interaction behaviours, which in turn improved achievement; the indirect effect ≈ was 0.87 (~ 33% of the total), indicating partial mediation. Effects were larger for girls and lower-income students. This study&#xa0;positions human–AI learning analytics including&#xa0;selective offloading&#xa0;and socially shared metacognition as pervasive computing signals that link family context to educational&#xa0;outcomes. It&#xa0;quantifies resilience as a moderator within an always-available AI tutoring ecosystem and offers design/governance implications for equitable AI deployment in schools.</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

The Moderating Effect of Resilience on the Relationship Between Parental Authority and Academic Performance in AI-Supported Learning Environments

  • 期刊论文

  • 发表日期：

    2026年3月28日

查看全文
  • Jing Qiaoqiao,
  • Chew Fong Peng,
  • Rusdi Bin Abd Rashid,
  • Sun Kun

摘要

The purpose of this study was to test whether student resilience moderates the relationship between perceived parental authority (authoritative, authoritarian, permissive) and academic performance in pervasive, AI-supported secondary-school learning environments, and whether human–AI interaction behaviours (shared metacognition, cognitive offloading) mediate these effects. The design of this study was a cross-sectional survey of 447 Grade 10–12 students (Guangdong, March 2025) in schools using an always-available, classroom AI tutor. Parenting styles (PAQ), resilience (school resilience scale), and AI-use constructs were measured; academic performance was self-reported (0–100) with robustness checks using quartile rank. We estimated ordinary least square (OLS) models with resilience interactions, a structural equation model (SEM) for mediation via shared metacognition and cognitive offloading, school-clustered SEs, and extensive sensitivity analyses. Authoritative parenting predicted higher grades (≈ + 2.1 points, p < 0.01); authoritarian predicted lower (≈ − 1.5, p < 0.05); permissive was null. Resilience strengthened the authoritative effect and buffered the authoritarian penalty (interaction terms ≈ + 1.0, p < 0.05). Resilience increased AI interaction behaviours, which in turn improved achievement; the indirect effect ≈ was 0.87 (~ 33% of the total), indicating partial mediation. Effects were larger for girls and lower-income students. This study positions human–AI learning analytics including selective offloading and socially shared metacognition as pervasive computing signals that link family context to educational outcomes. It quantifies resilience as a moderator within an always-available AI tutoring ecosystem and offers design/governance implications for equitable AI deployment in schools.