Cenare Cenarum wrote:
> I have never used STATA, therefore the question is-
> have I understood it correctly that in order to do a
> panel data analysis with binary and categorical
> outcome variables, it is better to use STATA than
> SPSS? The reasons for that are that STATA has logistic
> regression that taken into account that measurements
> are repeated and that STATA gives opportunity for more
> complex analysis compared to SPSS where basically all
> what one could do is to regress change scores for
> independent variables to change score (or value in the
> last wave) of the outcome variable. I am strugling
> through longitudinal data analysis books on my own and
> have no experience with this type of analysis,
> therefore I will appreciate any comments and advice
> (but please do take into account- I am neither
> statistician, no mathematician, just a PhD student in
> social sciences:)).
I cannot comment on SPSS, but, yes, Stata has a range of advanced tools for
panel analysis. These tools go far beyond of what you describe is available
in SPSS. You will found fixed effects panel regressions for metric and binary
outcomes, as well as for count data. There will be random effects models for
several other variants of the general linear model. And there will be some
other models which allow, you name it, more complex analyses. All these
models do indeed take into account that measurements are repeated.
The Stata command -help xt- will show you an overview on the tools that are
available in Stata.
Besides the common econometric books on panel analysis you may take a look
into "Kohler/Kreuter: Data analysis using Stata" usefull. Accidentally, I
know that the authors of this book had persons like you in mind while writing
the section on "regression models for panel data".
Many regards
Uli
--
[email protected]
+49 (030) 25491-361
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
