An Application of the Analytic Network Process (ANP) for University Selection Decisions

Abstract

Choosing to attend the right university for first degree study is one of the most complex real-life problems for some prospective students and their guardians. In addition, the outcome of the decision will affect not only the decision-makers but also their family, society and ultimately all sectors of the country. As such, selecting the universities is a critical decision that is just as important as making business decisions. From statistical data, the dropout rate after first year of university and a rate of lateral transfer from one faculty to another in public university are high. Therefore, students need to have a framework to assist them in achieving their lifetime goals. The research methodology begins with investigating and determining the Key Performance Indicators (KPI) influencing university selection decisions from relevant sources. Then, the process of synthesis using Nominal Group Technique (NGT) is implemented with small expert group and grouping them to each main factor/cluster with Affinity Diagram Technique (ADT). The university ranking with a systematic framework, called Analytical Network Process (ANP), is constructed. The Delphi method and focus group are applied to elicit experts’ opinion for identifying membership and relationship of the clusters/elements. Our objective is to test the ANP applicability with the process of selecting a university program. The paper proposes a framework to select universities for prospective students in order to achieve their different objectives and purposes. The framework consists of a series of steps that begin with the identification of the overall goal to achieve the desired result. Sensitivity analysis is also performed on multiple control criteria and sub-criteria. ANP is used to resolve this decision, rather than just the Analytic Hierarchy Process (AHP), which cannot accommodate the variety of interactions, dependencies and feedback between higher and lower level elements. According to priorities grouped under benefits, costs and risks within the control criteria, the results are based upon the relationships between clusters. Several scenarios are analyzed by varying different weights and ratings in the model to determine their effect on the results. Consequently, one can identify the most “satisfactory” university program on the basis of a number of both objective and subjective attributes and the proposed model is applied to the engineering discipline in Thailand.