Fig. 4. Student tracking and persistence vary as a function of school.

a Students with high education-associated polygenic scores are more likely to be tracked into advanced math in advantaged schools than in disadvantaged schools. Fitted probabilities of being tracked to Geometry or higher in the 9th grade, based on cumulative link logit model. School status measured by percent of students whose mother graduated from high school. See Supplementary Table 3 for full model results. b Students with low education-associated polygenic scores persist more in math in advantaged schools than in disadvantaged schools. Model-implied number of advancing steps from 9–to–12th-grade, based on Poisson model. At least 1 year of math beyond the 9th grade was compulsory in most U.S. states. See Supplementary Table 4 for full model results.