The role of school leadership on student achievement: Evidence from TIMSS2003

Abstract

Leadership, and especially head-teachers’ leadership, has been object of study since the late ‘60s, but the concept of leadership is neither unanimously defined, nor a consensus has been yet reached on its actual role and actual relevance within the school environment (Fullan, 2001; Sergiovanni, 2001; Harris, 2005). Good leadership can certainly contribute to school improvement by abetting the motivation, participation, and coordination of the teachers; recent studies have widened the range of action of school leadership research to the various organizational levels: school managers, department heads, coordinators, teachers (Goldhaber, 2002; Harris, 2004), and distributed leadership that could yield a higher impact on student achievement than what yet shown (Spillane et al., 2001, 2004). This dissertation takes its moves within the strand of research that identifies a significant role of leadership for student achievement (e.g. Edmonds, 1979; Cheng, 2002; Marzano, 2003) and tries to understand whether there are patterns of behavior of head-teachers that yield better results than others with respect to facilitating the student learning process and whether such patterns are consistent or replicable across countries. To address this question, the study uses the TIMSS2003 and investigates the relationship between head-teacher time allocation and school characteristics, student background, and student achievement in 18 countries. The model used in the empirical analysis is a three level Multilevel Model with random effects (evaluated using the R-Statistics software) that aims at evaluating the interaction effect between a particular school level variable (the time used by the head-teacher in managerial or leadership activities) and the explanatory variables describing school and student characteristics. What the study shows is that head-teacher specialization (either in management or in leadership) has negligible direct effect on student achievement. Most of all, however, head-teacher specialization is correlated to a lower impact of family SES on student achievement. Moreover, by investigating the impact of school management and school leadership on student achievement on students with different family background in terms of education, it is apparent that the high concentrations of school leadership are especially valuable for students of lower SES. On the other hand, the high concentrations of school management are most valuable for the students of higher SES. One possible explanation of these effects is that the attentiveness to the leadership process implies a deep involvement of the head-teacher in activities related to the modeling and tailoring of the educational process to the needs of the students. Such process has its highest payoffs on the students who come from disadvantaged situations and need special attentions in order to fully express their potential. On the other hand, the focus on the managerial side aims at rationalizing and making the best use of resources. This approach has high payoffs on students of all extractions, but is specifically relevant for the students of higher SES who are possibly already quite independent and whose performance can improve autonomously by making use of the extra resources that the management can provide. The analysis replication of the analysis on a country-by-country level confirms the existence of the afore-mentioned effects. More specifically, the results of the analysis suggest that, in the majority of cases, the head-teacher specialization appears to be correlated to a significant reduction in the dependence of student results from their family socioeconomic status. The same effect can be identified for both Math and Science in most countries. Nonetheless, the identification of a specialization-effect does not say much about the reasons for its existence. One possible explanation is that head-teachers are professionals that try to use at its best the opportunities provided by the institutional setup of the school system. In the more decentralized school systems that leave to the schools responsibilities in terms of monetary sanctions/incentives (hiring and firing, salary upgrades…), the head-teachers would tend to make use of these opportunities and focus most on management activities. Vice versa, in more centralized school systems, which leave to the schools only responsibilities that do not involve a monetary side, the head-teachers would stress their roles as role-models, educators, and motivators for their staff and collaborators. Hence, the final part of the research investigates whether the effect of the declared head-teacher specialization appears to go in the same direction as it could be predicted by looking at the macro-level institutional characteristics of the school system. These data suggest that school leadership and school management do have an impact on student results. However, the variables that enter in the process of determining the head-teacher time allocation are too many for indicating any specific policy direction based on average country behaviors. Still, the specialization of head-teachers in leadership or management is related to significant turnouts in terms of reduced needs of the students to rely only on the family resources (family SES) for improving their performance. In policy terms, such results suggest the need of allowing for different managerial strategies that could exploit local knowledge leads to foster the system’s equity and excellence. The dissertation is organized in 4 chapter plus 2 appendixes. The first introductory chapter looks at the economic nature of the educational good, the importance of its dissemination, and what are some of the possible interaction schemes among the system actors. The second chapter dives in the concepts of school leadership and management by looking at how it has evolved in the past 40 years. Subsequently, it addresses the limitations of the studies that have tried to establish a link between school leadership and student achievement, suggests how these limits can be overcome by means of a more comprehensive definition of the concept and of more advanced statistical techniques. The third chapter presents the research project on the TIMSS2003 dataset, the operationalization of the variables, the model for the statistical analysis, and the results of the study. The fourth chapter further discusses the results by contextualizing them within the legal and operational frameworks of the analyzed educational systems, and it concludes by addressing the limitation of the study, the indications for further research, and the possible suggestions in terms of policy making. The first appendix presents in detail the statistics for all the countries under investigation. The second and last appendix shows the detailed results for the analytic models at aggregate level and reports the dispersion of the residuals for each model