Need help with logistic multivariat

[Linda Kash]

Hi Guys.
I am using 2 data the one is QoG data which is on the individual level and ESS data which is on the Country level. And as you know both of these datas are already coded. Namley there is no missings in the variables. The purpose of my work is to merge these two data and then analyse them.

In ESS data I have
1 dependent variable
1 indep variable
2 control variables

In Qog data i have
1 independent
1 control


However, when i go back to ESS data which is the main data and merge the two datas, i got a problem with the reg command (see below)

reg migration_neg hinctnta gndr age4 freehms education logsocic

No observations

Its says that i don't have observation. But when i summarize my variables, I have observations.

Whats wrong? 😑

[Carlo Lazzaro]

Linda:
some comments about your query:
1) if you're planning to run a mutiple logistic regression (that is, a logistivìc regerssion with one regressand and 2 or more predictors), you need -logit- or -logistic- instead of -regress- (that invokes an OLS). In addition, a multivariare regression implies two or more regressands (which is not your case, as per your description);
2) two hypotheses about your last point. Your variable might be in -string- format and/or you've missing values.

[Linda Kash]

Linda:
some comments about your query:G
1) if you're planning to run a mutiple logistic regression (that is, a logistivìc regerssion with one regressand and 2 or more predictors), you need -logit- or -logistic- instead of -regress- (that invokes an OLS). In addition, a multivariare regression implies two or more regressands (which is not your case, as per your description);
2) two hypotheses about your last point. Your variable might be in -string- format and/or you've missing values.

Hi Carlo
I am actually following a instruction from a already made DO file. I dont know when my teacher do this, it works, But when i regress it is wrong.
How ever i have even tried to reg my variabels in ESS and Qog data separtly. It works well. But when i merges QOG and ESS i get no observations.

[Linda Kash]

I think I found the problem.
it is about the min and max. When I summarize all my variables I have som minimum which is -2 and some minimums which are 0
e.g education Min=0 Max=85.5

When I eliminate these variables which has minimum values of 0 and -2 I can regress.
So now, the question is can I do something that makes my minimum variables start from 1 and -2?

Thanks.

[Carlo Lazzaro]

Linda:
it's difficult tpo say without knowing your data.
For instance, you should be 100% sure that -2 is not the result of an error in data entry.