Neural change during spatial scanning predicts verbal reasoning transfer better than traditional performance-based assessments, and best prediction combines both¶

Claim · All claims
Evidence · 1 study (3 entries) · causal r2 · q2
Currency · newest study 2022 · not yet checked against its sources · not reviewed by a person · overdue for a recheck since 2025: tool and model accuracy changes quickly

Subclaims¶

q2 i? Each neural change predictor individually yielded significantly better model fit for predicting reasoning improvement than any performance-based assessment, including change on the same EFT task, PSAT, GPA, and course grades. → Cortes 2022 q2 i? Ensemble boosting analysis identified the two neural change variables as the most important predictors of reasoning improvement, ahead of ΔEFT and all traditional measures. → Cortes 2022 (2) q2 i? The strongest-fitting model combined the two neural change variables with three performance-based assessments. → Cortes 2022 (3)

Evidence¶

Cortes 2022¶

Cortes, R. A., Peterson, E. G., Kraemer, D. J. M., Kolvoord, R. A., Uttal, D. H., Dinh, N., Weinberger, A. B., Daker, R. J., Lyons, I. M., Goldman, D., & Green, A. E. (2022). Transfer from spatial education to verbal reasoning and prediction of transfer from learning-related neural change. Science Advances. https://doi.org/10.1126/sciadv.abo3555

q2 · i? · causal · r2

Multiple regression model comparisons in the neuroimaging sample; likelihood ratio tests compared neural change predictors against performance-based assessments including change on the same EFT task. Adding neural variables to the best performance-based combinations also significantly improved fit (all P ≤ 0.043).

"The two neural change predictor variables (ΔaIPS_EFT and ΔIPL_EFT) each individually yielded significantly better model fit for predicting ΔReasoning than any of the performance-based assessments in the full neuroimaging sample and in the sample of Geospatial students only [ΔEFT, ΔMRT, change in spatial habits of mind, PSAT, GPA, Geospatial course grade (Geospatial students only), and gender; likelihood ratio tests: all P < 0.001]."

Cortes 2022 (2)¶

Cortes, R. A., Peterson, E. G., Kraemer, D. J. M., Kolvoord, R. A., Uttal, D. H., Dinh, N., Weinberger, A. B., Daker, R. J., Lyons, I. M., Goldman, D., & Green, A. E. (2022). Transfer from spatial education to verbal reasoning and prediction of transfer from learning-related neural change. Science Advances. https://doi.org/10.1126/sciadv.abo3555

q2 · i? · causal · r2

Data-driven ensemble boosting analysis including ΔEFT, ΔMRT, spatial habits of mind change, PSAT, GPA, course grade, and gender; of performance-based variables, ΔEFT received the highest importance (16%). Analyses with course grades included converged on the same neural predictors.

"The prediction analysis identified the two neural change predictor variables as the variables that contributed most to prediction of ΔReasoning (table S33): ΔIPL_EFT (accounting for 38% of predictor "importance") and ΔaIPS_EFT (accounting for 28%)."

Cortes 2022 (3)¶

Cortes, R. A., Peterson, E. G., Kraemer, D. J. M., Kolvoord, R. A., Uttal, D. H., Dinh, N., Weinberger, A. B., Daker, R. J., Lyons, I. M., Goldman, D., & Green, A. E. (2022). Transfer from spatial education to verbal reasoning and prediction of transfer from learning-related neural change. Science Advances. https://doi.org/10.1126/sciadv.abo3555

q2 · i? · causal · r2

Post hoc model comparison: the best prediction combined neural and traditional measures, indicating neural change should bolster rather than replace traditional assessments. Adjusted R2 is not a standardized effect size, so impact is null.

"the strongest fit among all models was achieved by a model combining the two neural change variables (ΔaIPS_EFT and ΔIPL_EFT) with three performance-based assessments (ΔEFT, ΔMRT, and PSAT; adjusted R2 = 0.39; table S53)."

Discussion¶

Related Claims¶

• Curriculum-related neural changes during spatial scanning predict and mediate transfer to verbal reasoning — related
• Improvement in spatial scanning predicts and mediates the curriculum's transfer effect to verbal reasoning — related
• The spatial curriculum produced longitudinal neural changes during reasoning: increased aIPS activity, increased SpatialMap-to-DLPFC connectivity, and increased nonspatial-to-spatial representational similarity — related
• A spatial STEM curriculum yields far transfer to improved verbal deductive reasoning in high school students, including on nonspatial relation trials — related
• The spatial curriculum produced near transfer to improved spatial scanning and increased self-reported spatial habits of mind, but not to mental rotation — related