Pseudo-panels stata code

[Vanessa Barmpagianni]

Hi everyone,
I'm trying to construct a pseudo-panel dataset using catecorigal variables (dependent and independent) but I cannot find the stata code to do this. I have read basic papers related to this issue, but the most of these papers are reffered to linear models. My dependent variable is the self-rated health status (4 categoriries) and the independent variables are mainly demographic socio-economic characteristics grouped in categories as well (e.g. age, gender,education,income etc..). Can anyone help me??

Thank you in advance,
Vanessa (PhD Candidate)

[Carlo Lazzaro]

Vanessa:
welcome to this forum.
I assume that your health states are ordered: hence, I would take a look a -ologit-.
As an aside, the following Stata thread can be helpful https://www.stata.com/statalist/arch.../msg01305.html
Eventually, the following (quite expensive) textbook on panel data analysis (https://link.springer.com/chapter/10...no-access=true) included a chapter dedicated to the topic you're interested in, that you can also access at http://www.uio.no/studier/emner/sv/o.../PDAppl_17.pdf.

[Vanessa Barmpagianni]

Thank you for your quick response Mr. Lazzaro. Yes, the health status is an oredred variable. I have already run an ologit model and the results are consistent with the literature (actually, I investigate the relationship between income and health during economic crisis) , but I was thinking of differentiating my work from the rest by using and another method. The above links are very helpful as concerns the theoretical background, however I cannot find the code stata for pseudo-panels in order to do it in a right way.

[Carlo Lazzaro]

Vanessa:
as far as I know, there's no such a code.
Hpwever, you may want to type -searh pseudopanels- from within Stata and see the obtained entries.

[Richard Williams]

If you are with an academic institution, check for the availability of Carlo's suggested book, "The Econometrics of Panel Data." Mine gives me free online access to the whole book, which is cheaper than paying $379 for it.

If you want to differentiate your work, consider some of the alternatives to ologit, such as those provided by SSC programs like gologit2 and oglm.

But, the choice of method should be based on substantive concerns, not just a desire to impress people! So, if pseudo panels or other ordinal methods seem justified, go ahead and use them. But, doing something different may take some work and there are a lot of options out there, so think through them first.

You have probably already done that, of course. But, I'm reminded of two of my peers in grad school. I suggested to one that LISREL seemed appropriate for her problem, and she said "I am writing a substantive dissertation, not a methodological one." I told another student that LISREL seemed unnecessary for their problem, and he said that he wanted to impress potential employers and use it anyway. My own take is that you should use the method that is best for your problem, and distinctions like substantive vs methodological are generally nonsensical.

[Richard Williams]

The article "Pseudo‑panel methods and an example of application to Household Wealth data" says it used Stata:

https://www.insee.fr/en/statistiques...uillerm_EN.pdf

It makes reference to an article that contains Stata code. But the article is in French! If you can't find it or make any sense out of it, maybe you could contact the author directly.

If you google "pseudo panel stata" you get various other links. In particular, posts 10 and 12 in this thread suggest Stata code, but I don't know if it is correct.

https://www.statalist.org/forums/for...s-pseudo-panel

Finally, this thread also discusses pseudo panels, and some express skepticism about using them.

https://www.statalist.org/forums/for...s-section-data

[Vanessa Barmpagianni]

Fortunatelly, I have also free online access to the above mentioned book!! Concerning the proposed paper Mr. Williams, I have read it but, as in every paper, it doesn't gives exactly the appropriate commands for the used programme. Anyway,I would like to thank you both for your useful comments. I will take them into consideration.

[Yumeng Wang]

Dear all, have you find the Stata code of pseudo-panel methods now? I just send an email to [email protected] but it has been rejected. I humbly request you help me out.