parameter interpretation can be misleading for clm

[strengejacke]

Great. Just a heads-up: in clm() and other discrete choice models, when scale effects are included, which do not measure log-odds, coefficients in location and intercept parameters no longer measure log-odds directly, and exp(coef) are no longer odds ratios. So the following parameter interpretation can be misleading.
Parameter | Log-Odds | SE | 95% CI | z | p

Originally posted by @MrJerryTAO in easystats/insight#727 (comment)

[DrJerryTAO]

Thanks, @strengejacke. Intercepts do not measure log-odds per se either, if any scale parameter coefficients are present, as intercepts and location parameters are all in the numerator divided by exp(scale parameters linear combination) and this quotient is on the log-odds scale.