Inconclusive reinforcing effects between reading and mathematics: Scrutiny of the random-intercept cross-lagged panel model

Studies, e.g., using the random-intercept cross-lagged panel model (RI-CLPM), have indicated reinforcing prospective effects between reading and math achievement. However, effects in the RI-CLPM may be biased and spurious. Here, we analyzed data from a nationally representative sample of U.S. children (the Early Childhood Longitudinal Study - Kindergarten Class of 2010-11 (ECLS-K: 2011), N = 17,487) with the RI-CLPM as well as complementary models, e.g., multilevel models (MLM) of person-mean centered scores. The models indicated discrepant and inconclusive reinforcing and degrading prospective effects between reading and math achievement. Consequently, prospective effects between reading and math may have been spurious and conclusions of reinforcing effects would appear premature. It is important for researchers to bear in mind that correlations, including effects in the RI-CLPM, do not prove causality. For improved inference, researchers are recommended to triangulate and fit complementary models to their data in order to evaluate if effects from one model, e.g., the RI-CLPM, are corroborated or if they may be spurious.