I am trying to calculate the economic significance/ magnitude of the coefficient of my independent variable:
I am using a GEE model (
stata 14) to regress my dependent variable Y on my independent variable X and control variales.
The following GEE options are used:
Link: log Falimy: gamma Correlation (of my dependent variale):ar2
The coefficient of my independent variable x is
My questions are the following:
When I want to interpret the economic significance of my coefficient, do I procede the same way as a coefficient of an OLS regression? Hence, can I use the coefficient of my X to calculate the changes in my dependent variable Y to one-standard deviation increase of X? Or do I first need to alter the coefficient of my X in order to properly calculate the economic significance?
I found the following related postings which suggest that the interpretation of my X-coefficient through GEE is the same as a normal OLS, only that they should be considered as an "average" coefficient for the whole population: Interpretation of GEE coefficients, Interpreting ordinal GEE coefficients, or here Interpreting a longitudinal generalized estimating equations beta cofficients, which states that
The betas are very similar in interpretation to those from OLS, but for a population average. This suggests that to calculate the economic significance, I can use the coefficient of my X generated through GEE and treat it as an OLS coefficient?!
However, I have found the following from Ballinger (2004):
Because the log link function was specified, interpretation of the value of the parameter estimates requires that they be exponentiated by taking the log of the β coefficient estimates.
Does this mean, my first suggestion is false, and I first need to calculate the exponentiated value of my coefficient:
Thank you a lot in advance.
Ballinger (2004) Using Generalized Estimating Equations for Longitudinal Data Analysis, Organizational Research Methods, Vol 7, Issue 2