Hello,
I'm having some trouble specifying a regression using svyglm. Here's my current code:
dis<- subset(svydesign(
id = ~congl,
strata = ~strata,
check.strata = TRUE,
weights = ~fact, data = base_2022), bin == 1)
options(survey.lonely.psu="remove")
model_gamma <- svyglm(
income~ educ+task,design = dis,
family = Gamma(link = "log"))
summary(model_gamma)
My goal is to contrast the evolution of a specific coefficient over time. I have ten years of data from the same survey, so the complex survey design essentially remains consistent, with only the data=base_year parameter changing.
What technique should I use to evaluate the change in this coefficient over the years? I'm aware that handling standard errors with survey data can be challenging.
Thanks for your time and interest.
Have a nice week.