In a randomized experiment, personalized paths reduced learning time by 22.0% and improved post-test scores versus a fixed sequence¶

Claim · All claims
Evidence · 2 studies · 2 causal r2 · q3 · i2–i3
Currency · newest study 2026 · checked against its sources 2026-10-20 (automated, 2 of 2 entries) · not reviewed by a person

Subclaims¶

q3 i3 Students following personalized paths averaged 57.6 min versus 73.8 min for the fixed-sequence control group (d = 1.52), a 16.2 min (22.0%) saving. → Feng Z and Huang K 2026 q3 i2 After controlling for pre-test performance, the experimental group significantly outperformed the control group on the post-test (η² = 0.065). → Feng Z and Huang K 2026 (2)

Evidence¶

Feng Z and Huang K 2026¶

Feng Z and Huang K (2026) Bayesian cognitive diagnosis optimizes personalized learning paths via mediation of cognitive load and Hidden Markov Model state transitions. Front. Psychol. 17:1879982. https://doi.org/10.3389/fpsyg.2026.1879982

q3 · i3 · causal · r2

Randomized experiment with 120 undergraduates assigned 1:1 to personalized or fixed-order paths over the same probability-theory content. The personalized group's "average learning time ... 57.6 min" versus 73.8 min yielded d = 1.52 and a 22.0% time saving.

"the average learning time for students following personalized path 57.6 min (SD = 12.0), compared to 73.8 min (SD = 8.5) for the control group, a significant difference (t = 8.22, p < 0.001, d = 1.52). The time saving was 16.2 min (22.0%)."

Feng Z and Huang K 2026 (2)¶

Feng Z and Huang K (2026) Bayesian cognitive diagnosis optimizes personalized learning paths via mediation of cognitive load and Hidden Markov Model state transitions. Front. Psychol. 17:1879982. https://doi.org/10.3389/fpsyg.2026.1879982

q3 · i2 · causal · r2

ANCOVA on post-test scores (20 items) with pre-test as covariate in the same 120-participant randomized experiment; baseline pre-test scores did not differ (t = 0.52, p = 0.604). The printed effect size is "η2 = 0.065".

"the experimental group significantly outperformed the control group on the post-test after controlling for pre-test performance (F = 8.24, p = 0.005, η2 = 0.065)"

Discussion¶

Related Claims¶

• Intelligent tutoring systems improve achievement over non-individualised instruction — the general claim this finding bears on, with the studies behind it
• Cognitive load mediates the effect of personalized paths on learning outcomes, with motivation and self-efficacy as complementary mediators — related
• Personalized paths reduced self-reported cognitive load on all six adapted NASA-TLX dimensions with large effect sizes — related
• Shortest remediation paths respecting attribute prerequisites are shorter than random and full-coverage baseline paths — related
• Providing key features during case comparison reduces learning time — related
• In the 12-week quasi-experiment the experimental group's learning trajectory was significantly steeper than the control group's (group × time β = 0.52, 95% CI 0.31–0.73), with comparable baseline D-scores — related