.png){kind=icon}
.png){kind=icon}
Return
Reinforcement under uncertainty and expectancy-value dynamics: Effects of probabilistic feedback in an AI-mediated learning environment on motivation, emotion regulation, and learning engagement
.png){kind=icon}
PRE
AI
C
Y
DOI：10.1016/j.lmot.2026.102263.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Grounded in the integrated frameworks of Reinforcement Theory and Expectancy-Value Theory, this study addresses a gap in understanding the affective and behavioral impact of feedback schedules in technology-enhanced language learning. Although AI-mediated instruction is rapidly expanding, the role of probabilistic feedback, understood as reinforcement characterized by uncertainty, remains underexplored in EFL contexts, particularly in relation to motivation, emotion regulation, and learning engagement. To examine this issue, a quasi-experimental pretest-posttest design was conducted with 120 intermediate Chinese EFL learners recruited through purposive sampling and then randomly assigned to an Experimental Group (EG) receiving probabilistic reinforcement or a Control Group (CG) receiving fixed feedback during an AI-mediated reading intervention. Data were gathered using three validated instruments, and ANCOVA results showed that the EG obtained significantly higher adjusted posttest scores than the CG across motivation, emotion regulation, and learning engagement. The findings suggest that strategically designed probabilistic reinforcement can strengthen motivation, emotional adaptability, and sustained engagement more effectively than predictable feedback, offering an evidence-based principle for the pedagogical design of AI-driven language learning platforms.
Keywords:
Probabilistic feedback
Expectancy-Value Theory
Reinforcement Theory
Emotion regulation
Motivation
Learning engagement
AI-mediated instruction
EFL learners
Journal
L
IF:
1.8
Papers:
66
Citations:
1.4K
Organization
Cited Papers