Group preference aggregation rules: The results of a comparative analysis with in situ data
- PDF / 127,354 Bytes
- 7 Pages / 595 x 842 pts (A4) Page_size
- 98 Downloads / 214 Views
#2000 Operational Research Society Ltd. All rights reserved. 0160-5682/00 $15.00 www.stockton-press.co.uk/jors
Group preference aggregation rules: The results of a comparative analysis with in situ data MJ Read*1, JS Edwards2 and AE Gear3 1
University of Portsmouth, 2Aston University and 3University of Glamorgan
Committees are sometimes engaged in tasks concerned with establishing ranked priorities and rationing scarce resources. However there is no generally accepted approach to combining judgements from individual members of a committee to form a group consensus. There are issues concerned with the make-up of a committee, its size and history, and the procedures used to produce committee decisions. This article presents the results of an analysis of data from 28 separate committees engaged on an actual ranking task concerned with the selection of research proposals for public sector funding. This paper explores whether the ranked lists produced by the committees are in¯uenced by the position of a particular proposal in the agenda, the ®nancial value of the proposal, the scoring scale used by the committees, or an `out of line' rating by a single committee member. The results suggest that the last two factors can have a signi®cant effect on the ranked positions produced by the committees. Keywords: group support systems; public expenditure; ranking; prioritising; consensus
Introduction Numerous authors have addressed the problem of combining social choices on the part of a number of individuals into a form of group consensus for over 200 years. In this time no generally agreed approach has emerged, but the core issues are concerned with paradoxes of aggregation of individual preferences into a single group preference function coupled with more practical issues of effective group process, and quality of group output. This paper analyses the ranking of sets of research proposals into prioritised lists by a number of funding committees within a Research Funding Agency that provides funds and resources for scienti®c research. The research proposals are submitted by individual researchers in universities and higher education colleges, and are ranked by `expert' committees through a process of aggregating individual committee member ratings using a 5 point scoring scale. Those proposals that come towards the top of the ranked lists will be more likely to secure funding from the Funding Body, while those ranked towards the bottom will be unlikely to secure funding. To support this task of ranking the proposals, the Teamworker Group Support System (GSS), which uses individual wireless handsets to aid on-line scoring and ranking, was used by the committees; this system is described in detail in Gear and Read.1 For a review of
*Correspondence: Mr MJ Read, Department of Accounting & Management Science, Portsmouth Business School, Locksway Road, Milton, Southsea, Hants PO4 8JF
different types of GSS and applications see, for example, Finlay and Marples.2 It is the data collected from these committee sessions, through usi
Data Loading...
