Summarise posterior item strengths with credible intervals

This is a compact numerical companion to plot_strength_summary(). The reported strengths are relative: for Plackett–Luce fits, the default reporting scale has geometric mean one; for Bradley–Terry fits, a larger value means a higher chance of beating an item with a smaller value.

Usage

strength_summary(fit, credible_mass = 0.9)

Arguments

fit

A simple BT/PL or item-SBM btsbm_fit object.

credible_mass

Width of the central posterior interval, in (0, 1).

Value

A data frame with one row per item, ordered from larger to smaller posterior mean strength.