Using mathematical programming to solve large ranking problems
- PDF / 104,322 Bytes
- 7 Pages / 595 x 842 pts (A4) Page_size
- 51 Downloads / 200 Views
#2001 Operational Research Society Ltd. All rights reserved. 0160-5682/01 $15.00 www.palgrave-journals.com/jors
Using mathematical programming to solve large ranking problems ACB Tse* Chinese University of Hong Kong, Hong Kong This paper introduces a mathematical programming model that overcomes the major methodological problem of a large ranking task: respondent fatigue and deteriorated decision quality caused by an excessive number of objects to be ranked. The model was applied to the problem of ranking Marketing and International Business journals. There are more than 200 such journals, making direct ranking or rating very difficult, if not impossible. The result shows that the mathematical programming model uses very little information and yet can produce rankings that are in agreement with results obtained from direct ranking studies. Keywords: large ranking problem; mathematical programming
Introduction Many studies have found that there is a limit to human decision makers’ ability to process information and that decision quality may be affected if that limit is reached.1–4 Models of overstimulation5,6 have suggested that decision makers react negatively to environments characterised by high levels of crowdedness and information overload. According to the well-known model of Schroder et al,7 information processing decreases when a decision maker experiences information overload, and decision quality may deteriorate as a consequence.8,9 More recently, Hwang and Lin4 also found that both information diversity and information repetitiveness had a significant impact on prediction accuracy. Although there is evidence that a psychological state of information overload and overstimulation may lead to subsequent negative behaviour and is detrimental to decision quality, little work has been done to investigate possible methods that may be used to alleviate the problem or lessen the negative impact of information overload on decision quality. To fill in this gap in the literature, this paper reports a mathematical programming model that may be useful for the solution of a large ranking problem. Large ranking problems are commonplace in market research—particularly in conjoint and trade-off studies—and other areas of decision-making when respondent fatigue is a problem. Respondents typically feel the problem unmanageable when the number of alternatives given exceeds a certain threshold. In this research, the problem of ranking Marketing and International Business journals is used to illustrate the *Correspondence: ACB Tse, Department of Marketing, The Chinese University of Hong Kong, Shatin, Hong Kong. E-mail: [email protected]
usefulness of the mathematical programming model. The quality of academic journals has been a topic of intensive study because publication in reputable journals is a commonly used proxy for academic vigour and scholastic achievement. Many universities use publication in academic journals as the basis for tenure and problem decisions. However, there are usually a large number of journ
Data Loading...
