Examples of code for running aggregrated multilevel binomial regression models using R package "Rethinking."

Outcome variable: presence at amusement venues.

Intercepts-only model:

mi_amuse <- map2stan(
  alist(
    amusement_venue ~ binomial(n_amusement , p),
    logit(p) <- a_partic[p_factor],
    a_partic[p_factor] ~ dnorm(a , sigma),
    a ~ dnorm (0 , 2),
    sigma ~ dcauchy(0 , 1)),
  data = d, iter = 5000, chains = 4)


Weekend as predictor:

m_amusewknd <- map2stan(
  alist(
    amusement_venue ~ binomial(n_amusement , p),
    logit(p) <- a + a_partic[p_factor] + bw*weekend,
    a ~ dnorm (0 , 2),
    bw ~ dnorm (0 , 2),
    a_partic[p_factor] ~ dnorm(0, sigma_participant),
    sigma_participant ~ dcauchy(0 , 1)),
  data = d, iter = 5000, chains = 4)


Weekend and ALHB score as predictors:


m_amusealhb <- map2stan(
  alist(
    amusement_venue ~ binomial(n_amusement , p),
    logit(p) <- a + a_partic[p_factor] + bw*weekend + bl*alhb,
    a ~ dnorm (0 , 2),
    bw ~ dnorm (0 , 2),
    bl ~ dnorm (0 , 2),
    a_partic[p_factor] ~ dnorm(0, sigma_participant),
    sigma_participant ~ dcauchy(0 , 1)),
  data = d, iter = 5000, chains = 4)