["Journal of Computer Assisted Learning, Volume 42, Issue 4, August 2026. ", "\nABSTRACT\n\nBackground\nPrior studies in computer‐based feedback predominantly concentrated on feedback assigned by the program to learners, in which case learners only received feedback passively, without consideration for their preferences or autonomy. It remains unclear whether allowing learners to choose feedback types of varying complexity can improve learning.\n\n\nObjectives\nThe current study examined the effects of self‐selected feedback on emotions, motivation, and learning performance.\n\n\nMethods\nWe implemented a single‐factor between‐subjects design, involving 310 middle school students tasked with completing 10 multiple‐choice items on the topic of quadratic radicals. Learners either passively received program‐provided feedback with different complexities (Knowledge of Correct Response [KCR], Cue Feedback, and Explanation Feedback) or chose between any of these three feedback types after responses.\n\n\nResults and Conclusions\nThe results showed that: (a) learners in self‐selected feedback condition reported higher decisional autonomy than those in the program‐provided feedback conditions; (b) learners in self‐selected feedback condition experienced more positive emotions than those in the program‐provided KCR or cue feedback conditions, and reported stronger motivation than those in the program‐provided KCR; (c) learners in self‐selected feedback condition also outperformed those in the program‐provided KCR condition; (d) in the program‐provided feedback condition, more detailed information resulted in better learning performance (explanation > cue > KCR).\n\n\nImplications\nThese findings underscored the significance of allowing learners to choose feedback type to improve learning in computer‐based learning environments and provided empirical evidence to support more learner‐centered approaches to feedback design.\n\n"]