A survey-based priority elicitation protocol for community-based resource allocation decisions

Decision Analysis is increasingly being used to support resource allocation in communities, given its ability to represent the priorities of community members and support transparent resource allocation. Current attempts to elicit priorities in these interventions rely increasingly upon conducting surveys (face-to-face or online). Nevertheless, such preference elicitation initiatives may suffer if respondents do not clearly understand the questions being asked. Many protocols for priority elicitation currently used in Decision Analysis were originally designed to be employed by a decision analyst, who provides extensive support to a small number of decision makers in eliciting their judgments. However, such standard elicitation protocols may not be suitable for surveys, as the elicitation questions require a high level of understanding and a high cognitive effort from the respondents. Hence, in this paper, we suggest a new protocol for eliciting individual priorities for resource allocation decisions via either assisted or unassisted large-scale surveys, which elicits strict preference relations. We base this protocol on the Marketing research literature, which has dealt extensively with similar surveys. We adopt a Multi-Attribute Value Theory framework and design the protocol to avoid the range-insensitivity bias in multi-attribute choices. We assess the suitability of the widely employed swing weighting method for survey-based elicitation of priorities in comparison to the proposed protocol and find that swing weighting may not be suitable for resource allocation problems. We also suggest how the proposed protocol may improve the coherence of judgments elicited from the swing weighting method for survey-based priority elicitation.