LAUSR

A variable is considered ordinal when it exhibits an ordered categorical scale in which the distance between levels is unknown. Ordinal responses are used in many research fields and, for this reason, require proper statistical analysis. There are multiple methods for fitting ordinal regression (OR) models, as well as various software packages, mainly in R. In this study, we review and describe the R packages within the CRAN repository that can fit OR models through a systematic review adhered to the PRISMA statement. We identified 48 packages with diverse profiles in terms of specificity, modeling features, and model and link function versatility. Of these, 21 were designed for OR. A total of 34 packages use the frequentist approach, and 17 support mixed‐effects models. Nearly half incorporate variable selection methods, while seven can perform multivariate analysis, and eight support nonlinear predictors. The results also showed the cumulative logit model as the most recurrent model and the ordinal package as the most downloaded OR‐specific package. The most versatile package is VGAM, which includes all described links and models. To our knowledge, this is the first comprehensive review specifically focusing on OR models in R, so the findings of this study provide valuable insights. We intend to guide researchers towards the primary alternatives for fitting OR models in R, aiming to enhance their application in various research fields.