Multicategory logit models

Equations 3.4, 3.8, 3.10 to 3.13, 3.16 of Bilder & Loughin (2025): baseline-category logits and probabilities, one-at-a-time Wald intervals, proportional and non-proportional odds forms, the polr() parameterization, and category probabilities from cumulatives.

Usage

morie_multicategory_logit(
  bj0 = NA,
  bjs = NULL,
  xs = NULL,
  logits_2_to_j = NULL,
  cum_probs = NULL,
  j = NA,
  polr = FALSE,
  pi_hat = NA,
  var_pi = NA,
  z = 1.96
)

Arguments

bj0

Intercept.

bjs, xs

Slope and covariate vectors.

logits_2_to_j

Baseline logits for categories 2..J.

cum_probs

Cumulative probabilities excluding P(Y <= J) = 1.

j

Category index.

polr

If TRUE negate the slopes.

pi_hat, var_pi, z

Wald pieces.

Value

List with logit, probs, pi_j, wald (as supplied).

Examples

morie_multicategory_logit()
#> list()