27391
Stick or Switch? Factors That Influence Educational Placement over Time for Children with Autism.
A number of factors may influence education provision decisions for parents of children with autism, including advice from professionals, placement availability, primary school experience of inclusion and socio-economic status. The majority of studies exploring parental choice have taken a qualitative approach. Early child-, parent- and neighbourhood factors have not been systematically explored as predictors of educational support transition.
Objectives:
To use a longitudinal, community-based sample of children with autism, who were initially in mainstream primary schools, to investigate which factors predict later transition to specialist education, and investigate if later parent-rated school satisfaction significantly differs between the two groups (mainstream vs. specialist education).
Methods:
This observational study included 274 children with autism from two London boroughs, first seen at 4-9 years (Wave 1) and followed up 7 years later (age 11-15 years, Wave 2) when attending secondary school. Only children in mainstream school at Wave 1 were included in this analysis (n=157; 132 males, 25 females). Wave 1 measures included as predictors of secondary school placement were child IQ, child sex, parent-reported autism severity (Social Communication Questionnaire [SCQ]), parent and teacher-reported emotional/behavioural problems (Developmental Behaviour Checklist [DBC]), parental self-report of mental health (K10), parental education, ethnicity and neighbourhood deprivation. Secondary school placement was classified as mainstream (with/without support) or specialist provision. At Wave 2, parents completed the Engagement and Confidence Scale of the Wider Outcome Survey of Parents (WOSP-ECS; Humphrey et al., 2011). As IQ was an important predictor of placement, the association between each predictor variable and later school status was tested using logistic regression adjusted for IQ. Factors showing an association at p <0.1 were retained in the multivariate prediction model. The two parent-reported behaviour measures (SCQ and DBC) were highly correlated so an a priori decision was made to retain the parent-reported SCQ and the teacher-reported DBC in the fully adjusted model.
Results:
As expected, lower IQ predicted later school transition to a specialist setting. In addition, with IQ as a covariate, higher parent-reported SCQ, parent and teacher-reported DBC and white ethnicity also predicted specialist placement transition. However, only IQ remained significant in the full multivariate model, likely due in part to collinearity of the DBC and SCQ with IQ. Parents of children who had transferred to a specialist setting reported greater confidence with the school (WOSP-ECS M= 18.6, SD= 4.9) than those remaining in mainstream (WOSP-ECS M=16.4, SD= 5.3) (p=.02).
Conclusions:
In addition to IQ, results suggest that autism severity, emotional/behaviour problems and ethnicity may be important factors to consider regarding school placement for individuals with autism. These factors were fully significant in the simpler analyses and marginally significant in the multivariate analysis. In the current sample, it was difficult to disentangle ethnicity from locality and choice of education provision differences between the two boroughs. In the UK, there is still pressure for educational inclusion. The finding that parents of children who had transferred to specialist provision had more confidence in the school’s ability to support their child should be communicated to education policy makers.