Hello everyone,
To analyze my study I would like to do a Fixed effect regression - Panel Model.
For my study, all participants were asked to read different scenarios and to indicate their level of satisfaction, motivation, collaboration and assistance for each scenario on a Likert scale ranging from 1 to 7.
I also controlled for age, gender and employment status.
The dependent variables are motivation, collaboration, assistance and satisfaction.
The independent variables are the different scenarios and the control variables. I coded the scenario, gender and work status variables as dummy variables.
The aim is to examine the impact of the different scenarios on the dependent variables, using the first scenario as a reference.
Initially, I ran a Wilcoxon test, but to control for individual heterogeneity, I need to use a test that takes into account the within-subjects sampling design.
Challenge:
My dependent variables are ordinal.
My design is a within-subjects design.
My independent variables are categorical.
Question 1 : Can I perform a GEE ? Is it equivalent to a regression that takes into account the within subject design ?
Question 2 : I would like to do a Fixed effect regression. For that I need to go to Analyze > GLM > Repeated Measure.
However, I am a bit lost because I need to specify my whiten subject factor --> which for me is my 7 scenarios. Thus I indicated the Scenario with 7 level.
However, I would then like to test whether each scenario has an impact on one of my dependent variable. For instance : motivation.
But after I would like to know How can I test each scenario with the dependent variable ?Do I need to include the demographic variable or not ? And when do I need to include the dependent variable ?
THANK YOU
------------------------------
Emma Iannello
------------------------------