Racial and ethnic differences in the risks for reading difficulties across elementary school.

We investigated the early onset, over-time stability, and explanatory factors of racial and ethnic differences in the risks for reading difficulties including repeatedly across elementary school. We did so through analyses of a nationally representative cohort (N = 16,088). Fourteen percent, 17 %, and 14 % of Black, Hispanic, and Native American students compared to about 6 % and 8 % of White and Asian American students, respectively, displayed reading difficulties in kindergarten. Large differences continued to occur through fifth grade. About 26 % of Black, Hispanic, or Native American students compared to about 10 % of White or Asian American students ever displayed reading difficulties from first to fifth grade. Sixteen percent of Black students, 17 % of Hispanic students, and 12 % of Native American students displayed repeated reading difficulties. About 7 % and 4 % of White and Asian American students did so. Results from multivariable risk and resilience logistic regression models using the antecedent-opportunity-propensity theoretical framework indicated that other explanatory factors fully explained initially observed racial or ethnic differences in the risks for reading difficulties during elementary school. Particularly strong kindergarten predictors were the family's socioeconomic status (adjusted odds ratio range [aOR] of 0.75 to 0.86) and the student's reading achievement (aOR range of 0.16 to 0.56), mathematics achievement (aOR range of 0.33 to 0.50), working memory (aOR range of 0.70 to 0.79), and inhibitory control (aOR range of 0.71 to 0.83). Negative binomial regression models of repeated reading difficulties and robustness checks further supported the findings. Economic and educational policies to address racial and ethnic differences in the risks for reading difficulties including repeatedly over time may need to begin by kindergarten.