Dividing sample into subsamples in panel data

[Lubaba Prottoy]

Hello

I am working with a panel dataset of 3 year (2012, 2015 and 2018). I am studying the effect of health shock on consumption smoothing. I am conducting the analysis in an overall sample of households but want to look at the relationship for poor and nonpoor households separately as well. My concern is how best to go about it in the case of panel data. should I divide the sample in terms of the baseline data 2012?

Any suggestion is appreciated.

[Carlo Lazzaro]

Lubaba:
this may ba a research assumption that you shoud clarify in your research report/paper/whatever.

[Lubaba Prottoy]

Lubaba:
this may ba a research assumption that you shoud clarify in your research report/paper/whatever.

Thanks for your response. I wanted to know how subsample study in case panel data is considered in general? How would one go about it?

[Carlo Lazzaro]

Lubaba:
how subsample study (in addition full sample as baseline analysis) are considered depends on research field tribal rules.
My opinion (to be double-checked with colleagues in your research field) is that you can present some subsample analysis as sensitivity analysis.