Low Prior Knowledge students achieved significantly higher normalized learning gains than High Prior Knowledge students¶

Claim · All claims
Evidence · 1 study · causal r1 · q3 · i3 large
Currency · newest study 2019 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? The interaction between agency group and prior-knowledge level on normalized learning gain was not significant (p = .052). → Xingliang Chen & Antonija Mitrovic & Moffat Mathews 2019 q3 i3 There was a significant main effect of prior-knowledge level on normalized learning gain (ƞ2 = .144). → Xingliang Chen & Antonija Mitrovic & Moffat Mathews 2019

Evidence¶

Xingliang Chen & Antonija Mitrovic & Moffat Mathews 2019¶

Xingliang Chen & Antonija Mitrovic & Moffat Mathews. (2019). Investigating the Effect of Agency on Learning from Worked Examples, Erroneous Examples and Problem Solving. International Journal of Artificial Intelligence in Education (2019) 29:396–424. https://doi.org/10.1007/s40593-019-00179-x

q3 · i3 · causal · r1

Generalized mixed model with normalized learning gain as response, agency group and prior-knowledge level (LPK/HPK) as between-subject factors, in the same 40-participant SQL-Tutor study. The article reports the level main effect, F(1,35) = 5.87, p = .021, with effect size ƞ2 = .144.

"The interaction between group and level was not significant, F(1,35) = 4.062, p =. 0 5 2 , partial ƞ2 = .104. There was a significant main effect of the level, F(1,35) = 5.87, p =. 0 2 1 ,ƞ2 = .144. Therefore, the normalized gain of LPK students was significantly higher than the normalized gain of HPK students."

Discussion¶

Learner Variables¶

• Prior Knowledge — predictor: learners who differ on it differ in outcomes

Related Claims¶

• HPK students improved significantly only in the Low-Agency condition, with a large post-test advantage over HPK peers in High-Agency (d = .95) — related
• All training policies narrowed the prior-knowledge achievement gap from pretest to posttest, with BKT achieving the largest reduction (77.1%) — related
• High-ability students' learning behaves as a dominant α-process even at low pretest scores, while average-ability students show more γ-process behavior — related
• Under a dominant α-process, the normalized gain contains no pretest-score term and is uncorrelated with pretest score if α is uncorrelated with pretest score — reports the opposite
• When the γ-process (associative interaction of correct and incorrect knowledge) is considered, the normalized gain correlates positively with pretest score — reports the opposite
• Pre-instruction LCTSR scores correlate with normalized learning gains, most strongly for theoretical content (TUG-K r=0.59) and more weakly for descriptive content (DIRECT r=0.50) — related
• High- and Low-Agency conditions in SQL-Tutor do not differ in learning outcomes, though both improve significantly from pre- to post-test — related
• Students with low prior knowledge show significantly higher proportional learning gains than students with high prior knowledge when learning with MetaTutor — related