Experimental Design and Randomization Methods for Biomedical and Veterinary Research

Provides reproducible methods for experimental design and treatment allocation in biomedical, veterinary, agricultural, and clinical research. Includes simple, fixed-block, variable-block, stratified, stratified-block, cluster, matched-pair, restricted, minimization, and covariate-adaptive randomization, together with completely randomized, randomized-block, factorial, split-plot, Latin square, and crossover designs. Also provides allocation summaries, balance diagnostics, schedule export, and visualization. The methods are based on established principles of randomization and experimental design; see Rosenberger and Lachin (2015, ISBN:9781118742242) and Jones and Kenward (2014, ISBN:9781439861424).

ExpDesignR

Experimental Design and Randomization Methods for Biomedical and Veterinary Research

ExpDesignR provides reproducible tools for treatment allocation and experimental design. Version 1.0.0 establishes the first stable API for simple, blocked, stratified, cluster, matched-pair, restricted, and covariate-adaptive randomization, together with common experimental designs and allocation utilities.

Randomization

simple_randomization(100, c("Control", "Treatment"), seed = 123)
block_randomization(100, c("Control", "Treatment"), block_size = 4, seed = 123)
variable_block_randomization(100, c("Control", "Treatment"), c(4, 6, 8), seed = 123)
stratified_randomization(dat, "Sex", c("Control", "Treatment"), seed = 123)
stratified_block_randomization(dat, "Sex", c("Control", "Treatment"), 4, seed = 123)
cluster_randomization(paste0("Site_", 1:20), c("Control", "Treatment"), seed = 123)
matched_pair_randomization(dat, "Pair", c("Control", "Treatment"), seed = 123)
restricted_randomization(100, c("Control", "Treatment"), max_imbalance = 1, seed = 123)
minimization_randomization(dat, c("Sex", "Site"), seed = 123)
covariate_adaptive_randomization(dat, c("Sex", "Site"), seed = 123)

Experimental designs

completely_randomized_design(40, c("A", "B"), seed = 123)
randomized_block_design(40, c("A", "B"), block_size = 4, seed = 123)
factorial_design(list(Dose = c("Low", "High"), Diet = c("A", "B")), replicates = 3, seed = 123)
latin_square(LETTERS[1:4], seed = 123)
crossover_design(c("A", "B"), subjects = 20, periods = 2, seed = 123)

Utilities

allocation_summary(schedule)
plot_randomization(schedule)
export_schedule(schedule, tempfile(fileext = ".csv"))

The package uses established principles of randomization and experimental design; see Rosenberger and Lachin (2015) and Jones and Kenward (2014).