Dennis D. Leber

Allocating attribute-specific information-gathering resources to improve selection decisions

When collecting data to select an alternative from a finite set of alternatives that are described by multiple attributes, one must allocate effort to activities that provide information about the value of each attribute. This is a particularly relevant problem when the attribute values are estimated using experimental data. This paper discusses the problem of allocating an experimental budget amongst two attributes when the non-dominated decision alternatives form a concave efficient frontier. The results of a simulation study suggested allocation rules that take advantage of knowledge of the decision model and, when available, knowledge about the general shape of the frontier. These rules were compared to a default rule that equally allocated the experimental budget across the attributes. A proportional rule that allocated samples based on the value function weights performed well only in some cases; a more sophisticated step rule increased the frequency of correct selection across all weights.

Sample allocation for multiple attribute selection problems

Prior to making a multiple attribute selection decision, a decision-maker may collect information to estimate the value of each attribute for each alternative. In this work, we consider a fixed experimental sample budget and address the problem of how best to allocate this budget across three attributes when the attribute value estimates have a normally distributed measurement error. We illustrate that the allocation choice impacts the decision-makers ability to select the true best alternative. Through a simulation study we evaluate the performance of a common allocation approach of uniformly distributing the sample budget across the three attributes. We compare these results to the performance of several allocation rules that leverage the decision-makers preferences. We found that incorporating the decision-makers preferences into the allocation choice improves the probability of selecting the true best alternative.

Decision Analysis Methods for Selecting Consumer Services with Attribute Value Uncertainty

The basic risky decision is defined as a decision for which the outcome of an uncertain event, in addition to the alternative selected, defines the final consequence. Beyond the uncertain event, additional uncertainties can enter a decision including the uncertainty in the attribute values used to assess the decision consequences. When considering the selection of consumer products and services, formal and informal reviews of products and services are often used by consumers to estimate the level of satisfaction that will be received. When developing a decision model based on these data, attribute value uncertainty is often present and should be incorporated. In this chapter, we consider the uncertainty in the attribute values used to describe the possible consequences. We present several approaches to incorporate attribute value uncertainty into the decision analysis for choosing a roofing firm based on customer review data.