Fig. 3
From: Multiple processes independently predict motor learning
Conceptual illustrations of regression analyses used to examine motor learning. a Diagram showing how bivariate regression quantifies relationships between an individual predictor and motor learning without removing the variance shared with other potential predictors. b–d. Diagrams showing how multiple regression quantifies relationships between two (b), three (c) or four (d) predictors and motor learning. Regression coefficients estimate relationships from the independent and shared variance of each predictor, whereas semipartials estimate relationships from only the independent variance of each predictor. Light grey areas show portions of motor learning that cannot be attributed to a single predictor due to shared variance with other predictors. Dark grey areas show portions of motor learning that can be attributed to a single predictor after removing its shared variance