Comparing Data Envelopment Analysis and Human Decision Making Unit Rankings: A Survey Approach
-
Stanko Dimitrov
Abstract
In this paper we compare the ordinal rankings generated through Data Envelopment Analysis (DEA) methods to ordinal rankings generated by human decision makers. Through eliciting the total rank ordering for approximately 100 individuals on all of the four different datasets of Decision Making Units (DMUs), we compare the rankings generated by individuals to those generated by ten DEA methods. We observe that depending on the characteristics of the dataset one of the DEA methods performs better than the others in matching human decision makers.
Funding source: Natural Sciences and Engineering Research Council of Canada
Funding source: Mitacs
We would like to thank Brian Cozzarin for his helpful comments. We would also like to thank Tarun Khandelwal who, during his Mitacs Globalink summer project, developed the survey system and did some preliminary studies.
© 2015 by De Gruyter
Artikel in diesem Heft
- Frontmatter
- Proceedings of Flint International Statistics Conference Kettering University, June 24–28, 2014
- Urban Planning for Change: Data and Projections in City of Flint Master Plans (1920, 1960 & 2013)
- More Equal and Poorer, or Richer but More Unequal?
- How We Can Evaluate the Inequality in Flint
- Comparing Data Envelopment Analysis and Human Decision Making Unit Rankings: A Survey Approach
- Discrete Pareto Distributions
- Leverage Effect for Volatility with Generalized Laplace Error
Artikel in diesem Heft
- Frontmatter
- Proceedings of Flint International Statistics Conference Kettering University, June 24–28, 2014
- Urban Planning for Change: Data and Projections in City of Flint Master Plans (1920, 1960 & 2013)
- More Equal and Poorer, or Richer but More Unequal?
- How We Can Evaluate the Inequality in Flint
- Comparing Data Envelopment Analysis and Human Decision Making Unit Rankings: A Survey Approach
- Discrete Pareto Distributions
- Leverage Effect for Volatility with Generalized Laplace Error
