Bayesian Analysis for Multivariate Categorical Outcomes

Provides Bayesian methods for comparing groups on multiple binary outcomes. Includes basic tests using multivariate Bernoulli distributions, subgroup analysis via generalized linear models, and multilevel models for clustered data. For statistical underpinnings, see Kavelaars, Mulder, and Kaptein (2020) , Kavelaars, Mulder, and Kaptein (2024) , and Kavelaars, Mulder, and Kaptein (2023) . An interactive shiny app to perform sample size computations is available.

bmco

Bayesian methods for comparing groups on multiple binary outcomes.

Installation

# From CRAN (when accepted)
install.packages("bmco")

# Development version
# devtools::install_github("XynthiaKavelaars/bmco")

Quick Example

library(bmco)

# Generate data
set.seed(123)
data <- data.frame(
  treatment = rep(c("control", "drug"), each = 50),
  outcome1 = rbinom(100, 1, 0.5),
  outcome2 = rbinom(100, 1, 0.5)
)

# Analyze
result <- bmvb(
  data = data,
  grp = "treatment",
  grp_a = "control",
  grp_b = "drug",
  y_vars = c("outcome1", "outcome2"),
  n_it = 10000
)

print(result)

Functions

• bmvb(): Basic group comparison
• bglm(): Subgroup analysis
• bglmm(): Multilevel data

Getting Help

See vignette("introduction") for detailed examples.

Acknowledgements

The statistical underpinnings of this package were developed with financial support of a NWO (Dutch Research Council) research talent grant (no. 406.18.505) and the theoretical insights of Maurits Kaptein (Eindhoven University of Technology, The Netherlands) and Joris Mulder (Tilburg University, The Netherlands).