Why Put in This Much Effort?": How AI Availability Shapes Students’ Motivation in Introductory Programming

Synthesis: Tran, Harper & Price (2026) examine a pressing motivational paradox in contemporary computing education: the ready availability of AI tools that can complete programming assignments undermines students' willingness to invest effort in developing their own skills. Drawing on self-determination theory, the study identifies how the perception of AI as a 'shortcut' reduces autonomous motivation and fosters a transactional orientation toward learning. The findings resonate with existing work on Over-Reliance and Cognitive Offloading, suggesting that easy access to AI-generated code may erode the very persistence that produces deep learning.

The motivational paradox

Introductory programming is effort-intensive: learning to code requires sustained practice, debugging, and tolerance for failure. When an AI tool can complete an assignment in seconds, the perceived value of that effort collapses. Drawing on Self-Determination Theory, the authors argue this shifts students from autonomous motivation (learning for its own sake, driven by interest and mastery) toward a transactional orientation in which work is completed for the grade, not the skill. The availability of an effortless shortcut doesn't just make cheating easier — it makes the honest path feel pointless.

How AI availability shapes motivation

• Reduced autonomous motivation: the perception of AI as a shortcut lowers the intrinsic value of developing skills, because the goal (producing working code) can be reached without the effort that builds competence.
• Transactional learning orientation: students come to treat assignments as tasks to dispatch rather than opportunities to learn, optimizing for output over understanding.
• Persistence erosion: because Over-Reliance and easy access reduce the need for struggle, students miss the Productive Failure cycles that drive deep learning in programming.
• Equity concern: students who consciously resist AI assistance may fall behind peers who use it to complete work faster, complicating assessment fairness and the fairness of grading in Higher Education — a dynamic with direct implications for Academic Integrity.

Connection to the broader knowledge base

This paper sits at the intersection of Over-Reliance, Motivation, and Self-Regulated Learning. Its emphasis on why effort feels optional links to ai-availability-student-motivation-adjacent work on the cognitive costs of AI, to Agentic Literacy Debt: A Structural Problem the AI Literacy Field Has Not Yet Named (the cumulative skill deficit from habitual reliance), and to the socio-emotional side documented in Stuck in a Spiral": Shame and Guilt as Social Regulators of AI Use in Computing Education. For instructors, the motivational lens complements the tool-design and assessment-design interventions in Reducing AI Misuse: students are more likely to use AI productively when they have reasons — not just rules — to do the work themselves.

What this means for practice

• Learners. Do not read the availability of a shortcut as evidence that the work is not worth doing: six of the 13 interviewees questioned whether effort on an assignment was worthwhile once AI could produce it in minutes, and several strong students described the same debugging hours as demoralizing precisely because a faster route existed.
• Learners. Set explicit boundaries on AI use rather than relying on willpower: 8 of 13 reported feeling less accomplished after using AI, with heavier users describing the larger loss, and the students who escaped the confidence-eroding loop were those who kept use rare and bounded.
• Instructors. Assess the reasoning, not only the artifact: students' stated values were already pro-effort, so build checkpoints that make debugging, explanation, and revision visible instead of treating a working program as the evidence of learning — the redesign the study calls for belongs with authentic assessment and assessment validity, and it means framing goals AI cannot fulfill rather than treating output quality as the sole objective.
• Administrators. Make the AI rules concrete and leveled, with disclosure requirements attached to each level, because participants navigated a six-level policy in this course and still described taking shortcuts "sanctioned or otherwise" when a faster option was available.

Limitations

• The study rests on 13 semi-structured interviews with engineering majors in one introductory MATLAB course at a single large public research university, conducted in the final three weeks of one semester.
• Recruitment was voluntary and included a small extra-credit incentive toward a lab assignment; the authors acknowledge that self-selection may have biased the sample toward students with stronger opinions about AI.
• Per the study's interpretivist design, no inter-rater reliability was calculated, prior programming experience was categorized from participants' own interview descriptions rather than a survey measure, and the group is narrow beyond its size: 9 of 13 were White, three were first-generation students, and all were end-user programmers in engineering majors rather than computing majors.
• Every student was interviewed once, in the final weeks of the semester, and the protocol opened with a 60-second video demonstration of AI's programming capabilities before asking about motivation in general and then in relation to AI; the authors note this ordering may have primed AI thinking or created contrast effects, and that a single snapshot cannot show whether these patterns are stable orientations.

Citation

Keith Tran, Colton Harper, Thomas Price (2026). "Why Put in This Much Effort?": How AI Availability Shapes Students' Motivation in Introductory Programming.