“…Each model included a similar set of pre‐defined control variables to minimize the differences between schools, other than participation in FUTP60. Some of our previous work showed that these variables can partially explain the variability in the distributions of health‐related fitness in children, and therefore, warranted inclusion in our models. These included, school size (categorical variable; ≤100, 101‐300, 301‐500, and >500 students enrolled), school location (categorical variable; rural, town, suburb, and city), SES (continuous variable; % of students receiving free and reduced lunch), and PE class load (categorical variable; ≤90 minutes/week versus >90 minutes/week).…”