panel tobit model

The aim of this study was to evaluate the educational and research efficiency of 27 public universities in Iran in the academic years 2014-2015 and 2015-2016. This was done using non-parametric methods of data envelopment analysis and parametric random border analysis for the input variables (number of incoming students, number of faculty members, number of staff and budget) and output variables (specific income, number of students studying, number of graduates and conference papers). Such information was extracted from the budget bill, statistics and documents belonging to the Institute of Higher Education Research and Planning and the Islamic World Database website. The results indicated that there were inefficiencies in academic units and that the degree of these inefficiencies varied from university to university. In addition, there was an average increase in the efficiency of all comprehensive universities in 2016 compared to 2015. Furthermore, the data showed, while the average educational efficiency of selected universities has increased over time, the average research efficiency has decreased. The results of the data validity test obtained via the evaluation method of random border analysis indicated the heterogeneity of variance of the study data. The variance inflation test was conducted with the aim of investigating the problem of alignment between explanatory variables. The results showed, alignment between some explanatory variables, including faculty members and budget. In addition, the tensile effect of the estimated coefficients of some research inputs was not significant for the educational and research outputs of selected universities. Thus, universities should be efficient when making resource allocation decisions.