An impressive 39 percent of the variance in the test score difference PAE end −PAE pre was explained by a multiple regression on R 2 Lin and PAE pre , with both factors coming out as highly significant predictors.
← all excerpts
Lord's paradox in a continuous setting and a regression artifact in numerical cognition research.
1
—
—