estat gof not working after svy: ologit - goodness of fit test

[rubia valente]

I am trying to use "estat gof" after running "svy: ologit" model and I got an error message: "invalid subcommand gof"
The dataset is from LAPOP (Americas Barometer)

Code:

svyset upm [pw=weight1500], strata(estratopri)

svy: ologit occ age female ed married employed rural pardo black mestizo indigenous other colori sexi i.country
estat gof

After the results I get:
. estat gof
invalid subcommand gof
r(321);


Does "estat gof" not work with ologit? A reviewer suggested we estimate the BIC and AIC for our models, and since BIC/AIC don't work with survey data, I am trying to figure out another way to test goodness of fit, does anyone have another suggestion?

Any help would be much appreciated.

[Rich Goldstein]

the Homer-lemeshow test (not a good test anyway) does not work after -ologit- and also does not work after -svy-; if calibration is what you want, a search on statelist should find lots of discussions of calibration plots (comparison of predicted and observed values); the user-written -pmcalplot- may be useful as an approximation (the help file does not mention ordered outcomes but you can probably follow its advice re: continuous variables and get something useful; use -search- to locate and download