The Rank Ordered Logit command (rologit; superceded by cmrologit in Stata 16) does not have an -or- option to exponentiate coefficients. Why not? I've seen various sources that say it it is fine to exponentiate coefficients and treat them as odds ratios, but with rologit you have to do the exponentiation yourself. Is there some statistical reason that rologit won't do the exponentiation for you? Or is it just one of those things that Stata Corp never got around to programming?
Related Posts with Why no OR option in rologit
Log transforming variablesTwo questions related to log transforming variables. I understand that we would want to log transfo…
power analysis for modified poisson regressionHello, I am trying to determine the power of one my analyses. The outcome is binary and relative ri…
"Not sorted"Dear All, I have individual level panel data that includes spells, with a 6 month *follow up* period…
Comparing two datasets by two variablesHi, I am very new to stata. I have x1, y1, z variables in data1.dta and x2, y2, N in data2.dta. I a…
initial values not feasible melogitHello -- I am running an melogit on a 330,000 person nested in ~1900 neighborhoods in 47 countries. …
Subscribe to:
Post Comments (Atom)
0 Response to Why no OR option in rologit
Post a Comment