Multidimensional heterogeneous panel data models: Dimensional perspective on university rankings

Nowadays, with the development of technology and the increasing data stack in the digital environment, models with more than one dimension are needed. Multidimensional panel data models are becoming attractive to data analysts under the requirement of this demand. Researchers and institutions use these models day by day. In multidimensional panel data models, dimensions can be nested or not nested within each other. In this study, in addition to the multidimensional studies in the literature, heterogeneous panel data models are applied to multidimensional panel data models. These models, which take into account heterogeneity on the basis of individual units, offer a different framework to the study. Heterogeneous panel data model estimators are derived for multidimensional panel data models. These estimators were used to model success rankings in university entrance exams. Variables affecting success rankings in all dimensions, especially Turkish test score, were generally significant and their signs were negative as expected. In this context, this study with multidimensional heterogeneous panel data models makes a difference by explaining the economic knowledge in the literature. Keywords: Multidimensional Panel Data, Heterogeneous Panel, Cross Sectional Correlation, Univesity Rankings