Table 3.
Multigroup growth curve model results for academic self-concept, with initial levels of stereotype awareness (intercept) predicting initial levels of self-concept (intercept) for RQ3, initial levels of stereotype awareness (intercept) predicting changes (slope) in self-concept for RQ4, and changes (slope) in stereotype awareness predicting changes (slope) in self-concept for RQ5.
School track | Lowest track | Combined track | Higher track | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Parameter | b | SE | Est/SE | p | b | SE | Est/SE | p | b | SE | Est/SE | p |
Intercept of academic self-concept | ||||||||||||
RQ3: Intercept of stereotypes | −0.069 | 0.068 | −1.016 | 0.348 | −0.137 | 0.074 | −1.856 | 0.095 | −0.251 | 0.073 | −3.444 | 0.001 |
Slope of stereotypes | 0.005 | 0.111 | 0.041 | 0.967 | 0.075 | 0.076 | 0.99 | 0.322 | 0.048 | 0.102 | 0.467 | 0.640 |
Gender | 0.112 | 0.034 | 3.309 | 0.001 | 0.066 | 0.041 | 1.606 | 0.108 | 0.042 | 0.039 | 1.06 | 0.289 |
Socioeconomic background | 0.024 | 0.020 | 1.198 | 0.231 | 0.009 | 0.019 | 0.482 | 0.630 | 0.028 | 0.019 | 1.478 | 0.140 |
Migration background | −0.027 | 0.044 | −0.624 | 0.533 | −0.040 | 0.130 | −0.304 | 0.761 | 0.050 | 0.049 | 1.036 | 0.300 |
Changes in academic self-concept (slope) | ||||||||||||
RQ4: Intercept of stereotypes | 0.018 | 0.085 | 0.211 | 0.833 | 0.042 | 0.081 | 0.516 | 0.833 | 0.019 | 0.068 | 0.282 | 0.833 |
RQ5: Slope of stereotypes | −0.104 | 0.152 | −0.683 | 0.636 | −0.102 | 0.086 | −1.186 | 0.424 | −0.182 | 0.109 | −1.674 | 0.424 |
Gender | −0.073 | 0.047 | −1.533 | 0.125 | 0.030 | 0.042 | 0.715 | 0.474 | 0.118 | 0.041 | 2.900 | 0.004 |
Socioeconomic background | −0.020 | 0.024 | −0.838 | 0.402 | 0.011 | 0.022 | 0.504 | 0.614 | −0.035 | 0.025 | −1.414 | 0.157 |
Migration background | 0.041 | 0.053 | 0.777 | 0.437 | 0.185 | 0.145 | 1.282 | 0.200 | −0.014 | 0.045 | −0.309 | 0.757 |
RQ = research question. Gender: 0 = female, 1 = male; Migration background: 0 = without, 1 = with; Statistically significant p-values (p < 0.05) are printed in bold. The Benjamini-Hochberg correction was applied to all p-values related to a research question (separately for each research question).