OBJECTIVES: Moral decision-making does not contain straightforward trade-offs. While Discrete Choice Models (DCMs) struggle to model and explicitly examine morality’s complex and subtle nature in human decision-making, the moral dimension of decisions is rarely considered to advance the choice modelling field further. This paper explores the potential of using both DCMs and Machine Learning (ML), i.e., data-driven, methods and the integration thereof for moral choice analysis in healthcare and beyond.

METHODS: An interdisciplinary literature search across four databases – Pubmed, Scopus, Web of Science, and ArXiv – was conducted to gather articles. Based on the Preferred Reporting Items for Systematic and Meta-analyses (PRISMA) guideline, studies were screened for eligibility on inclusion criteria and extracted attributes from eligible articles.

RESULTS: Of the 6,281 articles, we included 273 in the review. DCMs with utility maximization-based specifications were used in morally salient decision contexts. There has been a move to improve the behavioural realism of the models. Some explicitly considered morality in choice behaviours. The prediction power, optimisation algorithms, and ability to handle unstructured data were considered the main capabilities of ML models. However, none were developed for the analysis of moral choice behaviours.

CONCLUSIONS: Data-driven DCMs help obtain more insights into moral decision-making. We discussed that ML methods could (i) assist as an exploratory, assumption-free approach for model building and (ii) enhance conventional (theory-driven) DCMs by integration to improve the behavioural realism of the models. Both research endeavours can increase the attraction and applicability of DCMs for moral choice analysis in healthcare and beyond.