Supporting multiattribute decisions in scenario planning using a simple method based on ranks

Abstract

AbstractA neglected area of scenario planning is the provision of support for strategic decisions that involve multiple attributes. When the number of scenarios and attributes is large, conventional multiattribute decision analysis methods require the elicitation of a large number of values and weights, which can be demanding and time consuming for decision makers. This paper examines the effectiveness of using a simple approximation to the simple multiattribute rating method (SMART) that is based purely on the ranking of options and attributes. The method was tested on 250,000 simulated decision problems and was found to perform well when assessed on the basis of its hit rate (the percentage of times it identified the same best option as SMART) and the utility loss resulting from the approximation. In large problems, where simplifications are likely to be most useful, it outperformed an alternative approximation method, SMARTER, which is more complex to apply, and it was almost as effective as SMARTER on smaller problems.