Predicting Community Opposition to Inclusion in Schools: The Role of Social Dominance, Contact, Intergroup Anxiety, and Economic Conservatism

ABSTRACT

This study addresses community members’ attitudes toward inclusion, the practice of including students with disabilities in regular education classroom settings. Participants in Study 1 were 271 community adults, completing measures of prior contact with people with disabilities, social dominance orientation (SDO), economic conservatism, intergroup anxiety, prejudice, and opposition to inclusion. Results from structural equation modeling indicated that amount of intergroup anxiety predicted opposition to inclusion via the mediator, prejudice toward people with disabilities, and that amount of prior contact indirectly predicted prejudice toward people with disabilities through intergroup anxiety. SDO positively predicted both intergroup anxiety and prejudice in the model, with prejudice also mediating between SDO and opposition to inclusion. Both SDO and economic conservatism failed to exhibit direct predictive relationships with opposition to inclusion. Participants in Study 2 were 161 community adults. Contact was shown to exert an indirect effect on prejudice via intergroup anxiety, whereas intergroup anxiety impacted inclusive attitudes via prejudice. SDO exerted both direct and indirect (via prejudice) effects on opposition to inclusion.

Keywords:

• disabilities
• economic conservatism
• inclusion
• prejudice
• social dominance orientation

Notes

1. We recognize that the terms mentally handicapped and psychiatric patients are not consistent with current use of people-first language. However, since these terms were used to describe particular groups in the Duckitt (2006) study, we decided to maintain these terms when referring to that study.

2. Once again, scores on “prior contact” were recoded such that higher scores represented greater amounts of previous contact with people with disabilities.

3. Only one measured variable was used as an indicator of contact quantity.

4. The χ² goodness of fit test was statistically significant, χ²(76) = 151.457, p < .001, consistent with poor fit. Because of the impact of sample size on the χ² statistic, we chose not to use this index as a measure of fit.

5. The χ² goodness of fit test again was statistically significant, χ²(82) = 157.297, p < .001.

6. In both the first and second alternative models, prior contact failed to exert a significant (ps > .50) direct effect on opposition to inclusion.

7. The correlation of –.34 between SDO and prior contact in our structural model supports this interpretation.

8. The decision to remove items from the SDO and Opposition to Inclusion scales (used in Study 1) reflected an effort to reduce the length of the surveys being administered (as the questions in the second study were administered alongside a political items associated with an unrelated study).

9. Contact was also allowed to directly predict opposition to inclusion in the model and was found to be a nonsignificant predictor.