Bayesian Meta-Regression

Usage

meta_reg(
  data,
  studyvar,
  yi = NULL,
  vi = NULL,
  mods,
  event_ctrl = NULL,
  event_int = NULL,
  mean_ctrl = NULL,
  mean_int = NULL,
  sd_ctrl = NULL,
  sd_int = NULL,
  n_ctrl = NULL,
  n_int = NULL,
  likelihood = c("binomial", "gaussian", "poisson"),
  model_type = c("random_effect", "common_effect"),
  stage = c("two_stage", "one_stage"),
  center = TRUE,
  scale = FALSE,
  small_sample = c("none", "t_approx", "hjsk"),
  mu_prior = NULL,
  tau_prior = NULL,
  gamma_prior = NULL,
  beta_prior = NULL,
  beta_priors = NULL,
  custom_model = NULL,
  custom_data = NULL,
  chains = 4,
  iter_warmup = 1000,
  iter_sampling = 1000,
  adapt_delta = 0.95,
  seed = 1234,
  ...
)

Arguments

data: A data frame with one row per study.
studyvar: Character. Column name of the study identifier.
yi, vi: Character. Column names of pre-computed effect sizes and their sampling variances (two-stage only).
mods: One-sided formula specifying moderators (e.g. ~ age + dose).
event_ctrl, event_int: Character. Column names of event counts for binomial / Poisson likelihoods.
mean_ctrl, mean_int, sd_ctrl, sd_int: Character. Column names of arm means and SDs for the Gaussian likelihood.
n_ctrl, n_int: Character. Column names of arm sample sizes.
likelihood: Character. One of "binomial", "gaussian", "poisson".
model_type: Character. "random_effect" (default) or "common_effect".
stage: Character. "two_stage" (default) or "one_stage".
center: Logical. Mean-centre continuous moderators. Default TRUE.
scale: Logical. Scale continuous moderators to unit SD. Default FALSE.
small_sample: Character. Small-sample adjustment for two-stage models: "none", "t_approx", or "hjsk".
mu_prior: Prior on the intercept.
tau_prior: Prior on the between-study SD (random-effects models).
gamma_prior: Prior on the Gaussian arm-level intercept (one-stage only).
beta_prior: Default prior for every regression coefficient.
beta_priors: Named list of coefficient-specific priors, overriding beta_prior for those coefficients.
custom_model: Optional character scalar of Stan code overriding the generated program.
custom_data: Optional named list merged into the Stan data list.
chains, iter_warmup, iter_sampling, adapt_delta, seed: MCMC settings.
...: Passed to cmdstanr::sample().

Value

An object of class c("bayesma_metareg", "bayesma").