MultiResponseR - Analysis of Data from Multiple-Response Questionnaires
Provides a multiple-response chi-square framework for the
analysis of contingency tables arising from multiple-response
questionnaires, such as check-all-that-apply tasks, where
response options are crossed with a known grouping factor. The
framework accommodates within-block (e.g., within-subject)
designs, as commonly encountered in sensory evaluation. It
comprises a multiple-response chi-square test of homogeneity
with an associated dimensionality test, a multiple-response
Correspondence Analysis (CA), and per-cell multiple-response
hypergeometric tests. These methods extend their classical
counterparts by grounding inference in a null model that
properly accounts for the multiple-response nature of the data,
treating evaluations, rather than individual citations, as the
experimental units, yielding more statistically valid
conclusions than standard contingency table analyses. Details
may be found in Mahieu, Schlich, Visalli, and Cardot (2021).
<doi:10.1016/j.foodqual.2021.104256>.
