As a motivational example, we’ll develop a measure of the dimensions of Moral Foundations Theory for use with C-suite executives (CEOs, CFOs, CXOs). We’ll discuss the relevance of moral foundations in business and show they are distinct from existing constructs using AI-based comparisons of construct definitions between moral foundations and adjacent constructs.
We describe automated item generation (AIG) using generative AI, show how transformers can encode the AI generated items to check construct alignment with pseudo-discrimination, and how pseudo-factor analysis can yield a clean factor structure for empirical testing. Lastly, we’ll show how to fine-tune encoder models to develop back-up items for trialling.
As always, the ultimate discrimination we want is empirical item discrimination. Our early work showed pseudo discrimination and pseudo factor loadings each predict empirical parameters. As this step moves from the AI psychometric realm to conventional psychometrics, we will move to showing the empirical discrimination relation with the new measure as data arrives.
Next section
Do moral foundations matter in business?
Last section
Return home
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).