dk_cell_probs: Cell probabilities for the model with Don't Know

In guess: Adjust Estimates of Learning for Guessing

Cell probabilities for the model with Don't Know

Description

The multinomial cell probabilities implied by the latent class transition parameters, as in equation (2) of Cor and Sood. The nine latent transitions among guess (g), know (k) and don't know (d) are reduced to seven by the identifying assumption that people do not lose knowledge over the course of a short informative process: the know-to-guess and know-to-don't-know classes are identically zero and do not appear.

The seven remaining proportions sum to 1, and so do the nine cell probabilities returned here. Given those probabilities the parameters are recoverable in closed form – gamma / (1 - gamma) is x10 / x00 – with one over-identifying restriction left over, x1d / x0d = x10 / x00.

Usage

dk_cell_probs(gg, gk, gd, kk, dg, dk, dd, g1)

Arguments

gg	proportion guess -> guess
gk	proportion guess -> know
gd	proportion guess -> don't know
kk	proportion know -> know
dg	proportion don't know -> guess
dk	proportion don't know -> know
dd	proportion don't know -> don't know
g1	probability a guess is correct (gamma)

Value

numeric vector of length 9, ordered x00, x01, x0d, x10, x11, x1d, xd0, xd1, xdd

guess documentation built on Aug. 3, 2026, 9:07 a.m.