NUMERICAL VS CARDINAL MEASUREMENTS IN MULTIATTRIBUTE DECISION-MAKING - HOW EXACT IS ENOUGH

Multiattribute decision making can involve consideration of both quant itative and qualitative measures of criteria attaintment, Some decisio n support systems (decision aids) to help multiattribute decision maki ng quantify value functions, One of the most popular of these systems, multiattribute utility theory (MAUT), requires two types of input, De cision makers need to express the relative value of different attainme nt levels on each criterion, as well as express the relative importanc e of these criteria, Some systems (such as DECAID) require simple dire ct graphical input of value and criterion importance, Other systems (s uch as LOGICAL DECISION) use more complex means of expressing relative value, Either way, MAUT converts expressions of criterion importance into quantitative form. This study compares the relative stability of numerical results obtained through two decision support systems, DECAI D and LOGICAL DECISION (LD), used in the task of evaluation of multiat tribute alternatives, Additionally the relative stability of results w as measured by comparison with results obtained using an ordinal metho d, ZAPROS. ZAPROS is a decision support system for construction of a p artial order over the set of alternatives, It does not require convers ion of qualitative measures into quantitative form. The relations amon g alternatives are close to those based on ordinal dominance, The resu lts of experiments show that ordinal relationships between task parame ters are much more stable than those obtained from quantitative measur es. Resuits from DECAID and LD are much less coincident with each othe r than with results obtained through ZAPROS, Many inconsistencies were found in subject responses. It is concluded that more attention shoul d be given to the means of testing judgment consistency, and that in s ome cases, attempts to solve decision tasks through more ''exact'' jud gments of value function parameters may lead to erroneous results. (C) 1995 Academic Press, Inc.