Essays on the efficiency of schools and student achievement

This thesis consists of five essays that investigate the efficiency differences of schools, explain the differences in efficiency and student achievement and take a further step by looking inside schools with differing efficiency. In the studies presented in the essays, efficiency differences were estimated with various methods, both parametric and non-parametric, and at different points in time using both cross section and panel data. The use of panel data, in particular, brings new possibilities to the analysis and understanding of efficiency differences. The stability of the efficiency estimations was also tested with several models. Information on this matter is a prerequisite if efficiency differences are to be studied further. The causes of inefficiency were examined in this thesis with both quantitative and qualitative methods. The use of both of these methods broadened the scope of earlier studies and provided useful information on their advantages and disadvantages in analysing causes of efficiency. The application of different approaches also provided a richer description and understanding of the case schools, their processes and factors affecting efficiency. Finnish general upper secondary schools provided a good empirical context to analyse efficiency differences between schools and factors affecting student achievement, because the schooling is fairly short, lasting on average for three years, information is available on the earlier school achievement of the students as well as their family background, and most importantly, it concludes with a nation-wide matriculation examination. The empirical analyses in this thesis are based on register data on Finnish general upper secondary schools, their students and students’ parents in 1991, 1990-1998 and 2000- 2004. The register data are supplemented with survey data from 73 schools in 1995 and 27 semi-structured interviews of principals and teachers in 9 schools in 1999. The results of the studies show that there were efficiency differences between Finnish general upper secondary schools. The size of the differences depended on the method and model used for calculating the efficiency. The rankings of schools based on efficiency scores were also fairly unstable. Concerning organizational practices, very few of them were related to efficiency. The results of case studies showed that schools with caring views about the students, professional staff relations, participative management and decision making, and curriculum work which was perceived as a way to develop school and own work were more efficient. In addition to investigating efficiency differences, the studies in this thesis examined the influence of the comprehensive school grade point average, family background, school resources, length of studies and decentralization of test taking on performance in the matriculation examination with different methods and data and at different points in time. The results provided causal information on the effect of changes in school resources on student achievement and they showed that student achievement was not affected by the changes in resources. New evidence was also provided on the effect of lengthening of studies and decentralization of test taking in the matriculation examination. According to results, schools with longer length of studies and higher rate of decentralization performed worse in the matriculation examination