Data Envelopment Analysis in University Rankings

Data Envelope Analysis in University Rankings 0 Introduction Data Envelope Analysis (DEA) is a way for an organization's decision unit (DMU) to generate performance indicators with multiple inputs and outputs. (Dyson, 2000, OR Insight Vol 13 Issue 4). The DEA considers each unit in a linear programming sequential order and selects the most suitable weight for it. In this survey, the DMU is a university, the output belongs to another category, ie the selection / hypothesis of one dataset "completion rate" 12 The list of the top 100 universities is the most advanced "computer Science "comes from.

Bhagavath, (2006) defines the data envelope analysis as follows. "Data envelope analysis is a linear programming problem that provides a way to calculate the level of service efficiency within an organizational group.The efficiency of the organization is based on what the group observed best.Haq et al. (2010) We are using DEA for a number of reasons.It can use multiple inputs and outputs in the model in the model.It does not need to specify the parametric function form of the production function.This makes it easier for the administrator Finally, in addition to certain scale of sense assumptions, it is also possible to analyze non-profit organizations.This study compared the other non-parametric and parametric methods Because of many advantages, we use DEA.

Using Data Envelopment Analysis (DEA) (another technique is Free Disposal Hull, FDH) masks all other methods of nonparametric classes. Charnes, Cooper, and Rhodes (1978) introduced this technology and named it today. Data envelope analysis techniques use linear programming to generate segmented envelopes on data points. Although this method is widely used for research on technical efficiency, it has the disadvantage that randomness is not taken into measurement efficiency. Furthermore, the envelope curve does not exist everywhere. Our focus in this study is parametric technology.

A paper on nonparametric and econometric models that can successfully combine data envelope analysis is described in Auditbert et al. It is (2003). Together with data envelope analysis and Tobit model, they infer the social and healthy determinants of the efficiency of cotton farmers in the northern part of Cote d'Ivoire. As representatives of family health, they use high density malaria parasites in the blood of individuals. They use a two-step process; first, they use data envelope analysis to derive relative technical efficiency values, and then what they think they will affect efficiency And compare these efficiency scores. The variable "high density malaria parasite in the blood" entered the model at the second stage. Their findings show that malaria greatly reduces farmers' technical efficiency. They further concluded that the strength of infection is more important than its existence.