Fixed Effects regression in SPSS 22

[het kippenhok]

CONTENTS DELETED

The author has deleted this message.

[Art Kendall]

I  do not have access to my documentation, etc. to check syntax suggestions.  However, profitability is a continuous construct and coarsening it to a dichotomy throws away a lot of information.

[David Marso]

"...but in my SPSS Statistics 22 I can't find the Analyze>Mixed Models option in this version..."
I suspect you don't have the Advanced Statistics Module installed.  That is where you would find MIXED.
--

het kippenhok wrote

Dear all,
I'm totally new to statistics and doing research on whether profitability influences companies' graphical use and for that I have gathered data for 23 companies over 10 years. For all company years I noted whether they were profitable or not and the amount of graphs they included in their annual report. Now I want to know whether they included more/less graphs when they are profitable compared to when they make a loss. I think I need to run a fixed effects regression where the dependent variable Y is the number of graphs and the formula will be as follows:
 Yit = α + β*D + t + λ
α = constant
β = effect of profit
D = Dummy-variable for profit
t = time fixed effect
λ = industry fixed effect
Does that seem plausible given what I want to research?

I've found tutorials with (I assume other versions of) SPSS executing a regression like that but in my SPSS Statistics 22 I can't find the Analyze>Mixed Models option in this version. Is there another way to do this? Or is it nog possible in this SPSS version?
Furthermore, can anybody explain to me how I start? This might sound silly, but I don't have a clue as to in what format I should put my data (e.g. wide/long, a new variable column for every year or not, etc).

Thanks in advance for taking the time to help me out, much appreciated!
Mieke

[het kippenhok]

Thank you, that has helped a lot already, I have made a request for installation of the advanced module :). Any tips on how to enter the data?

I don't know whether to create a new variable for each year of observations and include every company's obtained data under that 'year variable' (wide view?) or to see each year of observations per company as a new observation of the 'regular variables' (profit&number of graphs) and include it under those (long view?).

Thanks a lot!!

[David Marso]

The MIXED procedure assumes data are in LONG format.

het kippenhok wrote

Thank you, that has helped a lot already, I have made a request for installation of the advanced module :). Any tips on how to enter the data?

I don't know whether to create a new variable for each year of observations and include every company's obtained data under that 'year variable' (wide view?) or to see each year of observations per company as a new observation of the 'regular variables' (profit&number of graphs) and include it under those (long view?).

Thanks a lot!!