TheMentalist
08-25-2011, 04:34 PM
Hello everybody,
I'm running some regressions on the European Social Survey (ESS) dataset and I have some doubts on how to introduce some kind of temporal dimension since this is not a panel survey but just a cross-sectional one.
My question is: does it make sense to regress a variable which is present only in the last one of the 4 waves of the ESS, on some other variables which instead are present over all the waves?
My idea is to try to evaluate the impact of (for ex.) being married between 2002-2008 (i.e. the 4 waves period) on being unemployed just in 2008 (the 4th and last wave).
Any suggestion or remark is warmly welcome, thank you in advance.
I'm running some regressions on the European Social Survey (ESS) dataset and I have some doubts on how to introduce some kind of temporal dimension since this is not a panel survey but just a cross-sectional one.
My question is: does it make sense to regress a variable which is present only in the last one of the 4 waves of the ESS, on some other variables which instead are present over all the waves?
My idea is to try to evaluate the impact of (for ex.) being married between 2002-2008 (i.e. the 4 waves period) on being unemployed just in 2008 (the 4th and last wave).
Any suggestion or remark is warmly welcome, thank you in advance.