Modeling the development of English reading enjoyment in AI-mediated IDLE: A complex dynamic systems theory perspective

As artificial intelligence (AI) becomes increasingly integrated into English learning, AI-supported informal digital environments have emerged as an important space for learners' engagement with English reading. However, little is known about how English reading enjoyment develops over time in these contexts. Drawing on Complex Dynamic Systems Theory, this study adopted an explanatory sequential design to examine the dynamic development of English reading enjoyment in AI-mediated informal reading. Quantitative data were collected from 362 college students three times over one semester, with approximately 4-week intervals, and analyzed using Latent Growth Curve Modeling (LGCM). In addition, thematic analysis of learners' reflection journals was conducted to interpret the quantitative results and examine the developmental trajectory of reading enjoyment in AI-mediated informal digital learning of English (AI-IDLE) contexts. The unconditional LGCM results showed a general increase in English reading enjoyment across the three measurement occasions, together with significant individual differences in both initial levels and rates of change. The conditional model showed that AI-IDLE engagement significantly predicted learners' initial enjoyment and later development. Qualitative findings further suggested that learners' AI-mediated reading experiences involved fluctuating affective-strategic reading. AI support helped reduce reading frustration and increase learners' willingness to continue reading, although some learners also reported concerns about over-reliance. The findings suggest that AI-mediated reading activities should not only provide adaptive cognitive support but also promote learners' strategic and emotionally sustainable AI use. This study contributes to longitudinal research on positive emotions in AI-IDLE and affective development in technology-enhanced language learning.