On the size of the model selection problem in a conjoint measurement experiment
Year: 2026
Type: preprint
Abstract: We examine how the number of models respecting marginality in model selection using Maximum Likelihood Conjoint Measurement (MLCM) varies with the number of dimensions tested and explain the logic for determining this number. The results are related to Dedekind numbers, which increase super-exponentially, and, thus, suggest a limit to the number of dimensions that can be practically tested using the MLCM paradigm.
Authors: Kenneth Knoblauch, Frederic Devinck
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7225007763