General linear model or mixed linear model

#1
I am doing my thesis for my master degree. I am looking at determinants of profitability for commercial banks.
Overall, I have Return on Asset as a dependent variable. As independent variables I have chosen Leverage, credit risk, liquidity risk, size effect, gdp, inflation, interest rate.
What is more my sample size is as follows: 24 banks, for 10 years 2004-2014. I am doing unbalanced data, as I have some data missed for some banks.
I am adopting panel model with fixed effect from each bank.
The problem is I don't know how to make fixed effect model in spss.
I have tried linear mixed model and general linear model. My results are not right I guess.
I hope for your help!
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
Not an SPSS user, but for clarification - you are just trying to run a longitudinal model in SPSS? You do not want to control for random effects of Banks, correct?


That doesn't seem like it should be to hard to search for, "SPSS" "Longitudinal Data" or "Panel Data"
 
#3
Not an SPSS user, but for clarification - you are just trying to run a longitudinal model in SPSS? You do not want to control for random effects of Banks, correct?


That doesn't seem like it should be to hard to search for, "SPSS" "Longitudinal Data" or "Panel Data"
As I understood from the theory panel model used either with random effect or fixed effect. what is more, according to the previous studies, in the panel model the individual effect from each bank can affect the robustness of results. thus individual effect of each bank can be fixed. i have watched several videos regarding the fixed effect model on youtube, but they use Stata to make it. but I have spss to run my regression, so I need to make panel model with fixed effect on it.