Evaluation of efficiencies in the education system using TIMSS data with two-stage bootstrap data envelopment analysis

Inefficient resource use and mismanagement at the school level have serious negative impacts on human resource development and labor market outcomes. Hence, in order to achieve greater efficiency in transforming school inputs into higher learning outcomes, it is necessary to complement the efficient management of physical as well as human resources. Therefore, this thesis aims to measure the effectiveness of education systems in a global context and to investigate the scope and consequences of inefficiency at the cross-country level in terms of resource and human management dimensions. In this context, only the part of the TIMSS exam, which is the largest and most comprehensive international student achievement assessment study in the world, administered every four years to students at the 4th and 8th grade level, which constitutes the basis of education, was taken into consideration. Thus, it was aimed to identify gaps in the contextual factors affecting learning at the very beginning of the education level and to identify any areas of weakness. The data used in the study were based on students' achievement in mathematics and science, together with variables obtained from questionnaires completed by students, teachers and school principals. For this purpose, nonparametric resampling data envelopment analysis (bootstrap VZA) was used to estimate the relative aggregate efficiency scores of countries. The bootstrap approach is used here because it is an easy way to analyze the sensitivity of efficiency scores to sampling variations of the estimated frontier. Denmark, Germany, and Hungary were found to be relatively efficient in the use of education inputs that affect math and science achievement, respectively, and these regions were found to operate at near-optimal levels and to use input consumption at near-optimal levels to produce the desired constant output. The six countries with the lowest efficiency are Montenegro, United Arab Emirates, Bosnia and Herzegovina, respectively. In the second stage, technical efficiency was regressed and the effects of schools on school management variables were analyzed using a Tobit model. Tobit regression analysis revealed that resource utilization and management can increase the input-oriented efficiency of schools in particular countries. These results suggest that the ineffectiveness of schools in the countries under study is both a lack of resources specific to mathematics and science teaching and a lack of optimal use and management of these same resources.