Evaluate AI-use interventions for the specific process they target, from verification quality to reliance decisions and persistent learning¶

Strategy · All strategies
Evidence · 1 claim (1 mixed) · 1 study (1 associational), q3 · 0 of 1 report an effect size · 1 claim rests on one study

Description¶

The review recommends that educational interventions be evaluated for the specific process they target rather than with a single self-report measure of responsible use. A lateral-reading lesson should be assessed for verification strategy and separately for verification success; a confidence prompt may alter metacognitive monitoring; a forced-delay interface may change reliance decisions; and error-correction activities influence learning when they induce substantive processing. Interventions reducing inappropriate acceptance should also be tested for unintended rejection of correct assistance.

Design Implications¶

Context¶

Requirements¶

• Designs should adjudicate AI-output quality, capture verification initiation, code process quality, score success, record the accept-revise-reject decision, and evaluate that decision against AI quality

Constraints¶

• When learning is intended, immediate performance should be followed by unaided retention or transfer

Target Learners¶

• higher education students

Target Learning Goals¶

• effective verification
• calibrated reliance
• learning that persists without AI support

Affordances¶

Claims¶

• Students frequently accepted misleading AI recommendations (overreliance) while rarely rejecting correct ones (underreliance) [~W]

Related Strategies¶

• Embed verification routines and cognitive forcing functions into AI-supported learning tasks
• Assess fidelity separately for instructional components and include student application when evaluating strategy instruction

Examples¶

-

Key Sources¶

• Wei J and Shang Y (2026). Beyond checking: verification quality, reliance calibration, and learning in generative AI-assisted higher education. Front. Psychol. 17:1965371. https://doi.org/10.3389/fpsyg.2026.1965371