A multinomial logistic regression showed significant associations for the Moderately Attained profile at both time points and for the Resistor profile at T2 (also marginal significance at T1, with 0.05 < p < 0.10 ) At T2, students who were one standard deviation above the mean on negative math emotions were less than half as likely to be in the Moderately Attained or the Resistor profile, compared to their odds of being in the Devalued/Ambivalent profile.
At T1, a multinomial logistic regression showed a marginally significant association (0.05 < p < 0.10) between CCMI and membership in the Moderately Attained profile: increasing CCMI by one standard deviation doubled the odds of being in that profile compared to the Devalued/ Ambivalent profile.
At both time points, CCMI was associated with higher odds of being in the Resistor profile compared to the Devalued/Ambivalent profile, but this association did not reach statistical significance (possibly due to sample size).
Stereotyping Reported racial stereotyping was higher in magnet schools at both time points, approaching significance at T1 and reaching statistical significance at T2 (see Table 11 ).