underdisp: Diagnostics and Models for Underdispersed Count Data

Tools for detecting and modeling underdispersion in count data (conditional variance below the conditional mean), a phenomenon overlooked by the Poisson and negative binomial defaults. Provides a screening diagnostic that benchmarks at-risk dispersion against a zero-truncated Poisson; the continuous parameter binomial (CPB) regression and its zero-truncated variant, with an interpretable observation-specific bound and high-dimensional fixed-effects support; validated bootstrap (for coefficients) and profile-likelihood (for the dispersion parameter) inference; and quantities of interest including predicted probabilities and the implied ceiling. The likelihood is implemented in C++ for speed.

Documentation:

Downloads:

Linking:

Please use the canonical form https://CRAN.R-project.org/package=underdisp to link to this page.