I have a dataset (n=6000) of individuals belonging to three different ancestry groups (african (n=2000), native american (n=2000), european (n=2000)). I am studying how well a predictor (x) explains body mass index (y) among the different ancestry groups. To do this, I used two linear regression models:
Model 1: BMI ~ predictor(x) + covariates(age,sex,etc)
Model 2: BMI ~ covariates(age,sex,etc)
I then calculated the incremental-r2, which is how much the explained variance of BMI increases when I include the predictor in the model.
Incremental-r2: r2(Model1) - r2(Model2)
For each ancestry group I calculated the incremental-r2 of the predictor. Now I would like to know how to calculate a confidence interval for each incremental-r2 and perform a test to estimate if the incremental-r2 between the three ancestry groups is statistically different or not. Any idea on how to do this in R?