multivariate ordered logistic regression

[Witold Orlik]

Hello Stata users,
I was trying to run ologit command with a number of variables and got just a bunch of errors....

kz021: factor variables may not contain negative values
_st_ivarlist(): 3598 Stata returned error
rob__check_varnames(): - function returned error
robust_init_eq_indepvars(): - function returned error
moptimize_init_eq_indepvars(): - function returned error
Mopt_model(): - function returned error
<istmt>: - function returned error

So I wonder: is there any assumption regarding format of the dependent variable - could it be ordinal based on 5 categories?

Any other suggestions are welcome

Thank you if you are willing to help

Never run this analysis before

[Clyde Schechter]

Showing the output without the code that produced it generally does not provide enough information. Please repost with the actual command that led to this

The one error message that I can explain without seeing the code is "kz021: factor variables may not contain negative values." Evidently you have a variable kz01 which is discrete and which you entered into the regression as i.kz021. However, such variables are not allowed to have negative values. So there are two possibilities. Either you intend kz021 to have (some) negative values--in that case you need to recode the variable to all positive values. Alternatively, you think kz021 has only non-negative values but there is an error in the data. In that case, -browse if kz021 < 0- will show you the offending observations and then you can do whatever is needed to fix them.