Training-level test problem scores were only marginally associated with training treatment (p=0.0563), with the BWE+BPS group scoring lower early in training¶

Claim · All claims
Evidence · 1 study · causal r? · q2
Currency · newest study 2024 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? A mixed-effect regression across training-level test problems (2.4–6.4) found only a marginal association between treatment and problem scores. → Preya Shabrina 2024

Evidence¶

Preya Shabrina 2024¶

Preya Shabrina, Behrooz Mostafavi, Mark Abdelshiheed, Min Chi, Tiffany Barnes. (2024). Investigating the Impact of Backward Strategy Learning in a Logic Tutor: Aiding Subgoal Learning Towards Improved Problem Solving. Int J Artif Intell Educ 34:825–861. https://doi.org/10.1007/s40593-023-00338-1

q2 · i? · causal · r?

Mixed-effect regression with problem id as random effect and treatment as fixed effect on training-level test scores; “p-value=0.0563”. The article also notes T2 received significantly lower scores in problem 2.4 and lower average scores in 3.4 and 4.4 early in training.

"The result of the analysis gave a p-value=0.0563 which indicates that students’ problem scores in training level test problems were marginally impacted by the type of training treatment they received."

Discussion¶

Related Claims¶

• Backward worked examples alone did not change student behavior or performance: the BWE-only group behaved like the control group — related
• Students trained with both BWE and BPS derived subgoals during proof construction with significantly higher efficiency than the other groups — related
• Students trained with both backward worked examples and backward problem solving significantly outperformed control and BWE-only students on posttest problem scores — related
• The BWE+BPS group constructed significantly shorter solutions than the other groups in only two posttest problems — related