Return to search

An Engineering Decision Support System (EDSS) with alternative-criterion pair evaluations

An Engineering Decision Support System, EDSS, was developed using Bayesian
mathematics which incorporates knowledge and confidence components observed in
alternative-criterion pair decision making. The separation of knowledge and confidence
has been previously unaccounted for in decision-making methods. EDSS provides
decision support to individuals or teams which must make choices between alternatives
using alternative-criterion pair evaluations. Further, EDSS was instanciated into computer
software.
The EDSS decision support system was statistically tested using two variables, mechanical
experience of the participants and the use of a decision method, at two different levels and
in a replicated factorial experiment. The experiment consisted of teams of subjects solving
a simple mechanical design problem. Data from the experiment was collected for eighteen
different metrics in four categories. This research reports on each of eighteen metrics
using the hypothesis that the use of EDSS will show improvements in, or positive impact
on, the following four categories: the decision making productivity of idea processing, the
decision-making process, the perception of the decisions made by the decision makers,
and the ease of use of a computer decision support tool.
Statistical results of the experiment showed that EDSS successfully matched ad-hoc and Pugh's decision matrix performance for sixteen of the eighteen metrics and statistically exceeded the remaining two. These two metrics are, the conduction of more evaluations of alternative-criterion pairs ,and increased problem understanding.
This research also shows that a new alternative-criterion pair evaluation method has been successfully created that provides for:
- A separation of knowledge and confidence in the Belief Model of decision making.
- Decision results without complete evaluation of all alternative-criterion pairs.
- Aggregation of preferences from team members.
- A convenient means for investigating decision improvements. / Graduation date: 1997

Identiferoai:union.ndltd.org:ORGSU/oai:ir.library.oregonstate.edu:1957/34010
Date24 April 1997
CreatorsHerling, Derald E.
ContributorsUllman, David G.
Source SetsOregon State University
Languageen_US
Detected LanguageEnglish
TypeThesis/Dissertation

Page generated in 0.002 seconds