Panel data analysis using SPSS

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

RuudM123
Posts: 5
Joined: Wed Jun 13, 2012 9:58 am

Panel data analysis using SPSS

Postby RuudM123 » Wed Jun 13, 2012 12:25 pm

I would like to conduct a regression with 4 continuous variables. One dependent and 3 independent. Between 2 independent variables I would like to test the interaction moderating effect as well.

The difficulty that I don't understand is that my model is based on 36 cases in 4 different time periods. This means 36 airline carriers are studied for 2004, 2005, 2006 and 2007. For the regression this means I have a total of 144 observations (36 x 4) in one data sheet for which I'd like to run a regression?

Although I am reading things like the decision between ordinary least squares, generalized least squares, choosing between fixed or random effects. Can somebody help me and explain which analysis method in SPSS I need to use to run a regression, and if required how I should decide and interpret the fixed and random effects decisions.

Thanks in advance.
statman
Administrator
Posts: 2757
Joined: Tue Jun 12, 2007 12:08 pm
Location: Florida, USA

Re: Panel data analysis using SPSS

Postby statman » Wed Jun 13, 2012 1:05 pm

It looks to me that you have several analysis/hypotheses in mind or wanting to explain. Some models that come to mind for you to look into are: Simple regression, Analysis of covariance, one of the repeated measures models - To name a few
See the note below

NOTE: Please read the Posting Guidelines and always tell us your OS, the SPSS version and information about your study and data!

Statman
Statistical Services
RuudM123
Posts: 5
Joined: Wed Jun 13, 2012 9:58 am

Re: Panel data analysis using SPSS

Postby RuudM123 » Thu Jun 14, 2012 7:00 am

Thank you for answering, but I read a lot that for such panel dataset that I use a normal simple regression is not suitable.

Other studies makes a distinction between ordinary least squares and generalized least squares regressions. And then state that OLS is not suitable for panel data.
Next, a GLS model needs to run the analysis and then you need to choose between fixed and random effects...

Is a simple regression a ordinary least squares regression?
And which regression analysis models from SPSS are generalized least squares regression models?


Thanks in advance!

Who is online

Users browsing this forum: No registered users and 1 guest

cron