Performance-Learning Gap
Quick Answer
The performance-learning gap is the dissociation between a learner's assisted task performance and their unassisted independent capacity on the same class of task. AI assistance can raise visible output quality while underlying skill stalls or regresses, because the assistant performs the cognitive operations the task was designed to exercise. The gap becomes visible only when the scaffold is removed and the learner is reassessed alone.
Performance-Learning Gap
The performance-learning gap is the dissociation between a learner's assisted task performance and their unassisted independent capacity on the same class of task. It belongs to the family of measurement errors in learning assessment: assisted output is treated as evidence of skill acquisition when it is only evidence of joint human-AI output. The gap becomes visible only when the scaffold is removed and the learner is reassessed alone. The canonical demonstration is Bastani et al.'s high-school mathematics field experiment, where unrestricted GPT-4 access raised practice scores by 48% but produced a 17% drop on unassisted exam performance versus control (Bastani et al., PNAS 2025).
The term is diagnostic, not normative. Structured tutors that fade scaffolding can close the gap; unrestricted assistants that perform the germane cognitive work for the learner tend to widen it.