Claims

A machine learning model using 105 predictors predicts next-year standardized math achievement with out-of-sample prediction R-squared of .70–.78

Claim · All claims
Evidence · 1 study · associational r3 · q3
Currency · newest study 2024 · not yet checked against its sources · not reviewed by a person

Subclaims

Reading the evidence codes

q is the design tier, i the impact, n the sample, and the kind with its r the rigour judged within that kind: what each code means.

This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.

q3 i? The best machine learning model predicted future achievement scores well, with prediction R-squared values of .70, .73, .78, and .77 for the Year 5→6 through Year 8→9 transitions. → Lavelle-Hill 2024

Evidence

Lavelle-Hill 2024

Lavelle-Hill, R., Frenzel, A. C., Goetz, T., Lichtenfeld, S., Marsh, H. W., Pekrun, R., Sakaki, M., Smith, G., & Murayama, K. (2024). How the Predictors of Math Achievement Change Over Time: A Longitudinal Machine Learning Approach. Journal of Educational Psychology, 116(8), 1383–1403. https://doi.org/10.1037/edu0000863

q3 · i? · associational · r3

Next-year prediction analysis of PALMA data: Elastic Net and Random Forest models trained on 80% of each data set with five-fold cross-validation and evaluated on a 20% hold-out test set. The best model's out-of-sample prediction R-squared ranged from .70 to .78 across the four transitions.

"We found that the best machine learning model could predict future achievement scores well, with the prediction R-squared values: Year 5 → 6 = .70 (mean absolute error [MAE]9 = 34.99), Year 6 → 7 = .73 (MAE = 34.44), Year 7 → 8 = .78 (MAE = 32.74), and Year 8 → 9 = .77 (MAE = 35.97)."

Discussion

Page metadata
Field Value
type claim
id ml-model-predicts-next-year-math-achievement-r2-70-78
title A machine learning model using 105 predictors predicts next-year standardized math achievement with out-of-sample prediction R-squared of .70–.78
description A machine learning model using 105 predictors predicts next-year standardized math achievement with out-of-sample prediction R-squared of .70–.78
status review
evidence_strength moderate
newest_evidence 2024
generated by process:wiki-ingest on 2026-10-25
trust unverified — no one has confirmed this page's content
sources 1 — lavelle-hill-2024

Frontmatter is the page's machine-readable record — see the Schema & Guide.