AI feedback does not motivate by itself: a motivational-meaning model for university physical education

Artificial intelligence can now watch a student’s body move, measure every joint angle, and issue a corrective cue within seconds. In university physical education classes around the world, computer vision and markerless pose estimation are converting movement into scores, visual overlays, and natural-language prompts. The technology is impressive, and vendors promise that precise feedback will motivate students to learn faster and exercise more. But a new theoretical article argues that this promise rests on a hidden assumption that does not survive scrutiny: that technical precision determines how learners interpret what they receive. According to researchers Xiangyu Wang and Yongwei Wang of Capital Normal University in Beijing, writing in Frontiers in Psychology, an AI output has no fixed motivational function independently of its use and interpretation. The same validated recording and correction, presented privately as a tentative suggestion or publicly as a compulsory judgment, holds its technical content constant while its instructional meaning changes entirely.

The model the authors develop is a motivational-meaning model, and its central claim is deliberately provocative for a field racing to deploy AI in gyms and sport halls. Drawing on self-determination theory, feedback literacy, self-regulated learning, and motor-learning research, the model specifies how three conditions organize the event in which a technical output becomes a psychological one. Pedagogical framing concerns how the output’s purpose, reference standard, rationale, and authority are communicated. Social visibility concerns who can see the feedback, how it is compared, and what consequences follow. Embodied verification concerns the learner’s access to opportunities for comparing the AI output with felt movement, task effects, replay, and knowledgeable interpretation. These conditions answer different questions about the same event: what is this output for, who can judge me through it, and what evidence can I use to assess it?

The theoretical engine is cognitive evaluation theory, a mini-theory within self-determination theory. External events do not carry one invariant motivational effect; their functional significance depends on whether they are interpreted primarily as informational, supporting a sense of effectiveness while preserving volition, or as controlling, pressuring the person toward a prescribed response. The same broad category of event can have different effects, which is why rewards and feedback should not be classified by form alone. Applied to AI, this means a precise indication of failure without an attainable next step may make inadequacy clearer rather than support competence. Likewise, nominal choice need not support autonomy if all options remain unintelligible or if questioning the output is prohibited. A friendly teacher can still present an algorithmic score as non-negotiable, and a demanding lesson can preserve agency through clear rationale and meaningful challenge.

University physical education is a distinctive setting for this problem, and the model treats it as the focal configuration rather than a generic classroom. Unlike feedback on a private written task, movement feedback concerns a performing body. The learner has felt the movement, observed its consequences, and may have a pre-reflective sense of timing, balance, effort, or discomfort. Felt experience can be mistaken, especially for novices, but it remains relevant evidence that may agree or conflict with the system’s representation. Competence may be seen by classmates while it is still developing, errors can become socially comparable, and recent evidence links perceived ability exposure with evaluation anxiety and lower psychological well-being among university students. When AI makes performance more precisely visible, storable, and shareable, it can intensify either the usefulness or the evaluative consequences of feedback. AI does not create visibility, but persistent scores, replayable video, and rapid ranking can change its scale and consequences.

The model also introduces a provisional state called blocked translation, which describes a temporary inability to establish a defensible evidential basis for the next action after encountering conflicting information. Every feedback system can be wrong, and every learner can misperceive a movement, but AI makes this relation theoretically important because the output may combine high apparent precision with low interpretability. A learner can see a figure of 87 percent incorrect without knowing the confidence interval, normative sample, camera limitation, or criterion weighting. When the score conflicts with bodily sensation or task success, the feedback event becomes an epistemic problem: what should guide the next attempt? The discrepancy can resolve through productive recalibration, when replay and testing show that the revised movement improves both the task outcome and felt control, or through justified correction of the system, when camera occlusion, atypical body proportions, or assistive equipment produces a misleading classification. Blockage is not a third motivational meaning and does not require the absence of controlling pressure; a learner can be unsure what is warranted while feeling compelled to comply.

Perhaps the model’s most important temporal distinction concerns what remains after feedback is removed. The authors propose self-regulatory appropriation, a process in which monitoring and adjustment initially organized by external feedback become demonstrable without the same support. A student who can state that they will attend to the pressure shift before landing and use balance on the next attempt as evidence has acquired more than a score; the learner has a cue, a comparison process, and a decision rule that can travel to a later task. Controlling feedback can produce a different trajectory, in which a mandatory target improves acquisition while leaving action dependent on an external instruction. This possibility is consistent with consumer research showing that quantifying an activity can increase the amount performed while reducing enjoyment and subsequent engagement, because the activity comes to feel more like work. Motor-learning research adds a parallel caution: the guidance hypothesis proposes that frequent external feedback can guide practice while limiting the development of independent error detection, and experiments on reduced feedback frequency and fading show why performance during practice cannot stand in for retention.

The model distinguishes five outcome domains that can diverge dramatically. System engagement concerns interaction with the technology, such as consultation or requests for feedback. Acquisition performance concerns execution during supported practice. Retention and transfer concern persistence after a delay and adaptation to a changed task or setting. Affect and well-being form a fourth domain, with event-level anxiety or enjoyment distinguished from well-being assessed across a learning unit. Continued participation in voluntary physical activity beyond the instructional requirement is the fifth, distal domain. A learner may repeatedly consult a system yet practice poorly, or a public rank may increase effort and immediate performance while reducing enjoyment. Conversely, a learner may cease using the system because its monitoring function can be performed independently, not because the intervention failed. Lower system use may reflect successful independence, lack of access, or disengagement, and is uninterpretable without complementary evidence.

Seven proposition groups specify separable predictions, and the authors insist that each be tested as a distinct primary contrast rather than as a comprehensive supportive-AI package. With output content, reference information, and audience held constant, mastery-oriented rather than pressuring framing should increase informational appraisal and reduce controlling appraisal. With the same individual output, norm reference, and grading consequence, named disclosure to peers rather than private display should increase controlling appraisal, with the effect predicted to be larger at lower baseline perceived competence. When independently established output-criterion disagreement is presented, withholding an otherwise available verification opportunity should increase next-step evidence-selection uncertainty relative to providing that opportunity. For a valid cue describing an attainable discrepancy, adding an intelligible next-action explanation should increase subsequent competence satisfaction relative to a matched description of the discrepancy alone. And when technical information, verification resources, teacher contact, and nominal feedback choice are held constant, an effective right to challenge and request revision of an output should increase subsequent autonomy satisfaction while improving discrimination in reliance across valid and invalid outputs.

Calibrated trust is a central measurement concept, and it is not high average trust. A learner who accepts both sound and erroneous cues has high reliance but poor discrimination; a learner who rejects both has low reliance and similarly poor discrimination. A proper test should score each output against an independent criterion, obtain a probability judgment of its correctness, and record acceptance, checking, deferral, or rejection. Probability calibration can be evaluated by comparing estimated and observed correctness, while conditional reliance distinguishes acceptance of valid from invalid advice. The two should be reported separately because a learner may distrust a cue yet follow it under grading pressure. This comparison is especially important when an AI output has high apparent authority but low explainability: a generic human cue may be recognized as one person’s judgment, but a numerical output can look definitive even when its training data, thresholds, or confidence are invisible. Trust should therefore be treated as a calibrated, revisable judgment rather than an outcome to maximize.

The authors are careful to state what the model does not claim. It offers testable design principles rather than evidence that AI feedback improves learning or participation, and continued participation is treated as a distal hypothesis requiring longitudinal evidence with attention to prior activity, health, habit, identity, access, and opportunity. Reviews of the neighboring evidence support caution: a systematic review of digital formative assessment in physical education identified only fifteen eligible studies and just five with retention tests, and meta-analytic work suggests that the benefits of self-controlled practice and enhanced expectancies in motor learning have been substantially exaggerated by underpowered designs and reporting bias. The model’s value will depend on precise contrasts, independent measurement, and willingness to discard unnecessary constructs. Its practical implications are theory-based principles for testing: make the purpose, reference, and limits of an output intelligible; connect discrepancies to attainable adjustments; limit involuntary evaluative exposure; provide both evidence resources and a credible challenge route; and examine when prompts can be reduced without losing useful structure. AI feedback does not motivate by itself, the authors conclude, and the account specifies how that claim can be investigated without confusing supported performance, learner regulation, and activity beyond the course.

Subject of Research: AI feedback does not motivate by itself: a motivational-meaning model for university physical education

Article Title: AI feedback does not motivate by itself: a motivational-meaning model for university physical education

Article References: Wang, X., & Wang, Y. (2026). AI feedback does not motivate by itself: a motivational-meaning model for university physical education. Frontiers in Psychology, 17, Article 1951652. https://doi.org/10.3389/fpsyg.2026.1951652

Image Credits: AI Generated

DOI: 10.3389/fpsyg.2026.1951652

Keywords: feedback, does, motivate, itself, motivational-meaning, model, university, physical, education, scientific research