GLS Model suggested by Stata instead of OLS (time-series data)

[Denise Klueber]

Hello everyone,
this is my first post, please excuse any mistakes I might make when posting.

For my master thesis, I am working with time series data (N (109)>T (72)) which I want to use to find out which effect imposed sanctions (sanctions) have on violence against women (VAW, varible: ageallf) in the target state. I am fairly new to stata and quantitative analysis.
I intended to work with the OLS regression model, but once I entered the regression command to test my model, Stata automatically went with a GLS regression. I have read reasons for why GLS is used instead of OLS and understand the reasoning. But I don't understand how I can know my data is heteroskedastic if I can't test it without automatically using the GLS model. If I do tests after the regression, the results are based on the GLS regression and result in no heteroskedasticity.

Any help is appreciated.

Kind regards,
Denise

[George Ford]

show your commands and output

[Carlo Lazzaro]

Denise:
welcome to this forum.
Echoing George's wise advice to read and act on the FAQ to increase your chances of getting helpful replies, with a bit of a guess-work I read your post as follows:
1) you're dealing with a N>T panel dataset that shows a long T;
2) you coded -xtreg,re- and Stata, as expected, performed a GLS regression.

Futher details are needed from your side to reply more positively, that is another way to say "please help others help you out !". Thanks.