Adjust for variables in logistic regression

[Eirin Fraenkl]

Dear stata users,

I am fairly new to working with stata so I hope I can find some answers from experts in this forum!

I am currently working with a survey sample with approx. 19.000 observations, the data was collected online across different countries.
We want to do a binary logistic regression (dependent Variable coded as 0 and 1) with several independent variables (categorical and continous).
I would like to adjust the model for variables such as country etc.

Do I simply add the adjustment variables in the "logistic" command as independent variables in the model? Is there another way to adjust in a binary logistic regression?

Thank you in advance

[Carlo Lazzaro]

Eirin:
welcome to this forum.
Some comments about your query:
- use the -svy:- prefix (see -help svy-) if your data were obtained from a survey;
- the only way to add predictors is to put them in the right-hand side of the regression equation.
If your predictors are categorical variables and/or interactions, take a look at -fvvarlist- (see (-help fvvarlist-).

[Eirin Fraenkl]

Dear Carlo,

thank you for the quick response!

So would it be correct (after defining survey settings) to command

svy: logistic dreaming i.Finance i.Household i.Maritalstatus i.gender c.age

if my dependent variable is dreaming, my independent variables are Finance, Household and Maritalstatus and my adjustment variables are gender and age? (as an example)
Or do i need to specify the adjustment variables in another way?

Thank you

[Carlo Lazzaro]

Erin:
regardless my wonder about entering -dreaming- as a regressand (please consider it as an amateur's comment), your code (that you could have better reported using CODE delimiters, though. Please, see the FAQ on that. Thanks):

svy: logistic dreaming i.Finance i.Household i.Maritalstatus i.gender c.age

, your code makes sense: the right-hand side of your regression equation should contain both predictors and controls.
You also used the -fvvarlist- notation the right way (and that is laudable for a self-reported Stata newbie).

[Eirin Fraenkl]

Thank you very much for your help!