Skip to main content
. 2018 Jan 2;47(6):1151–1163. doi: 10.1007/s10964-017-0795-0

Table 2.

Logistic regression predicting adolescents’ engagement in ethnic harassment from negative attitudes toward immigrants

Model 1a Model 2b Model 3c
b SE OR b SE OR b SE OR
Intercept −2.37** 0.15 0.09 −4.83** 0.56 0.01 −8.24** 1.60 0.00
SES −0.33* 0.15 0.72 −0.34* 0.15 0.71 −0.31 0.16 0.73
Adolescents’ negative attitudes toward immigrants 0.97** 0.20 2.64 0.78** 0.22 2.18
Friends’ negative attitudes toward immigrants 0.10 0.29 1.10
Class negative attitudes toward immigrants 1.58* 0.77 4.86

a Model 1 summary: AIC = 4509.59, BIC = 4514.33, level-2 variance component = 0.37, SE = 0.19, p = .06

b Model 2 summary: AIC = 4549.17, BIC = 4553.90, level-2 variance component = 0.26, SE = 0.18, p = .16

c Model 3 summary: We initially estimated the level-2 variance component in the model; however, the Hessian matrix was not positive definite due to lack of variance between classrooms. Therefore, we set the level-2 variance component at zero and re-estimated the model, for which AIC = 450.94, BIC = 474.26.

*p< .05; **p< .001