Comparing Formation Methods for Value Models for the NASA Artemis Human Landing System

There are many ways to form value models, but few comparisons of value model formation methods in literature. This paper provides a comparison of two methods that could be used to form a value model: multiple linear regression (a statistical technique using explanatory variables) and the general monotone model, GeMM (a statistical modeling technique designed to deal with “messy” data). Two value models for the NASA Human Landing System (HLS) for the Artemis Program are formed using the two methods, each based on the same system attributes. The ordinal consistency of the value models is analyzed. This research will formally highlight the challenges faced by value modelers in forming a value model and will provide a comparison between two proposed methods, identifying potential benefits, assumptions, and restrictions for each.