Search reports, ask a question, or jump to a page.
The problem
Intervention: Comparison of criterion-related validity across three informant sources (child self-reports, teacher reports, principal reports) of school safety against outcomes in three domains: children's experience, institutional functioning, and externally assessed academic achievement
Who it was tested with: 395,961 children in Grade 4 across 12,955 schools in 63 education systems; 27,100 teacher records.
Experience from people who deliver programmes: stories, case studies and lessons learned. Valuable for understanding how and why, and for spotting what to test next, but not a measure of effect.
"Students were randomly split within schools so that no student contributed both a predictor and an outcome; the split was replicated 50 times and averaged. Convergence was evaluated with a multitrait-multi-informant model, with ceiling effects and differential unreliability specified as rival explanations; informant contributions were estimated by an exact relative-importance (LMG) variance decomposition." This design permits causal inference by preventing overlap between predictor and outcome data and tests competing explanations for informant differences, qualifying as experimental evidence.
The three informant scores jointly explained only 1-13% of between-school variance, indicating other unmeasured factors are important. Study is cross-sectional and cannot establish causality. Results depend on TIMSS survey design and items, which may not fully capture all dimensions of school safety.
Across countries and designs
1 study · 12 countries
