Assessing the unit of measurement for school climate through psychometric and outcome analyses of the school climate survey

This study examines the appropriate unit of theory of school climate while simultaneously conducting a psychometric analysis of a modification of the School Climate Survey for an elementary school population. A conceptualization of school climate as an individual-level property unique to each school participant is compared to the idea that climate is a characteristic of the school with school participants having the role of informants. These two theories are compared by testing their differing implications. The School Climate Survey is found to have a stable and identical factor structure both within and between schools, ratings by different respondent groups are found to be moderately correlated, and average school climate within each school is found to predict a statistically significant amount of the between-school variation in children's academic achievement and cognitive functioning. Differences between individual raters within each school are not found to have a meaningful relationship to child outcomes. It is concluded that the School Climate Survey is a valid and useful measure of school climate and the school-level theory is found to have the strongest support.