An Empirical Comparison of Discrete Choice Models for Mining Stakeholder Analysis

This research sought to facilitate improved stakeholder analysis in mining by providing further insights into the preferences of local community members using discrete choice theory. While recent research has demonstrated the usefulness of discrete choice theory in mining stakeholder analysis, no previous work has examined which discrete choice model (DCM) is most suitable. This paper provides a research note on a case study in a mining community that was performed to compare three DCMs. After a thorough examination of the benefits and deficiencies of all models, this study concludes that the conditional logit model stratified by questions is the most useful DCM for mining stakeholder analysis. The recommendation is based on the usefulness and accuracy (ability to match the survey data) of this model.