A closer-to-mastery task-selection constraint sharply reduces overpractice for the risk-averse strategy, with large Cohen's d effect sizes¶

Claim · All claims
Evidence · 1 study · design r2 · q2 · i3 large
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i3 Guiding the Minimize Worst Case Loss strategy toward skills closer to mastery reduced overpractice from 30.28 to 1.83 (equation-solving) and 28.93 to 3.22 (graph-interpretation), with Cohen's d of 1.87 and 1.86. → Haley Noh 2026

Evidence¶

Haley Noh 2026¶

Haley Noh, Aarna Chowdhary, Jeroen Ooge, Vincent Aleven, Conrad Borchers. (2026). Simulating Learners' Task-Selection Strategies and System Constraints in Mastery Learning. arXiv preprint. https://arxiv.org/abs/2605.21613 doi:10.48550/arxiv.2605.21613

q2 · i3 · design · r2

Simulation results (Figure 3) comparing constrained versus unconstrained task-selection for the Minimize Worst Case Loss strategy across both datasets. The printed effect sizes are "large respective Cohen's d effect sizes of 1.87 and 1.86". Constraints bounded rather than removed learner choice.

"when the risk-averse strategy (Minimize Worst Case Loss) was guided toward skills closer to mastery, the excess overpractice was sharply reduced from 30.28 to 1.83 times in equation-solving and 28.93 to 3.22 times in graph-interpretation relative to the unconstrained baseline, with large respective Cohen's d effect sizes of 1.87 and 1.86"

Discussion¶

Related Claims¶

• A prefer-harder problem-selection constraint reduces overpractice for the risk-averse strategy in the equation-solving dataset — related
• The risk-averse Minimize Worst Case Loss task-selection strategy produces far higher overpractice than other strategies in simulated mastery learning — related
• In simulated mastery learning, system constraints left most other task-selection strategies' overpractice stable while sharply reducing it under the outlier risk-averse strategy — a broader claim this one bears on