Abstract (Translated):
We propose a management tool to identify groups of public schools that have similar standards of performance and therefore need similar support and guidance.The student's result is measured with the "Prova Brasil" (PB) for mathematics and for Portuguese language, and with an indicator of the condition of being performing (PB) with a higher than expected age. The degree of inclusion measured by the rates of disapproval and dropping out of school. We defined as the school product the portion of the students' results that is not explained with their family and school environment, and we identified a partition in groups of schools with dissimilar patterns, but composed of similar schools.We implemented this methodology with information from the "Prova Brasil" and the School Census of the year 2013. The main empirical results are: i) it is possible to define a partition of the set of public schools in 9 groups according to the learning and the degree of inclusion of the students; ii) these groups are linearly separable; and iii) this partition is partially explained with unobserved variables common to all schools in a municipality.