- Adjusting for dose response for outcomes:
Recall also that we re-ran analyses restricting the exposed group to include only students whose teachers were categorized as “high fidelity”. These results also did not differ from the analyses where students of all teachers in the exposed condition were included.
- Adjusting for high fidelity for outcomes:
As stated in Section 3, we ran sensitivity analyses to explore whether results for any differences between the students exposed and unexposed to PAX were affected by:
1) excluding students in the schools that dropped out of the PAX RCT prior to PAX implementation;
2) excluding students in the exposed group whose teachers had not attended the PAX training in the 2011/12 school year.
We also ran a dose response analysis that used number of days since teacher training rather than comparing exposed to unexposed.
Furthermore, we ran an analysis restricting the exposed group to include only those students whose teachers were categorized as “high fidelity”.
As mentioned in Section 3, we also conducted a series of stratified analyses, stratifying students by sex, income quintile, mother’s age at first birth, school location (urban/rural), family receipt of income assistance, and family involvement with CFS. None of these analyses yielded statistically significant results for any of the outcomes.
- Outcomes Stratified by Child Characteristics