Abstract
While self?evaluation leads to valuable results in some schools, it appears that in other schools this is true only to a lesser extent or not at all. This raises the question of how differences in the results of self?evaluations can be explained. This study looks at to what extent the results of self?evaluation are determined by the way in which self?evaluation is conducted, by characteristics relating to the general functioning of the school and by the support which schools enjoy. One thousand seven hundred and eighty?six school principals and team members from 96 schools (primary and secondary) were surveyed by means of a written questionnaire. The data collected were then analysed using multi?level structural equation modelling. The results provide strong empirical evidence that ?attitude with regard to self?evaluation?, ?self?evaluation as a policy action? and ?self?evaluation as an act of research? are powerful predictors of the quality of self?evaluations