How do people aggregate value? An experiment with relative importance of criteria and relative goodness of alternatives as inputs

Abstract

The concept of importance of criteria is used as a central element in several decision making contexts, specifically in value aggregation, e.g. as an input to decision support tools. For example, in the analytic hierarchy process (AHP) decision makers are asked to estimate how much more important one criterion is than another. However, it is not clear how people understand aggregation models based on importance of criteria in decision making situations. The purpose of this descriptive study is to investigate if people find an aggregation model in simple value aggregation tasks which remind of the way AHP elicits the input. Further, the purpose is to investigate if people's tendency to find a model depends on their cognitive abilities. In an exploratory laboratory experiment, participants assessed which of two alternatives is the best, based on information about the importance of two criteria and how good the two alternatives are compared to each other with respect to these criteria. The results confirm that people are willing to use importance of criteria and goodness of alternatives as input in value aggregations and show three main models for aggregation. More participants with higher numeracy applied a clear model compared to those with lower numeracy. None of the identified models was one of AHP's models but one of them reminded of one of the ways input can be aggregated in the AHP. The three models identified in the experiment are based on lexicographic order, multiplication and a combination of multiplication and addition. How the results could be used in a prescriptive context is discussed in the paper.