Hi. I am using betaconv.ols and sigmaconv.t functions from REAT package to measure both beta and sigma income convergence within EU28 from 2005-2017. Everything works well, but I want to check the data for heteroskedasticity and autocorrelation. I have tried F test for heteroskedasticity (ols_test_f) and within the function the first argument is the model, which is an object of class lm. I did not use lm function, because it is already within betaconv.ols function. So I checked the source code for betaconv.ols which uses linear regression to calculate beta. I found this: conv_model <- lm (ln_growth ~ ln_initial)
I have applied this and created my own lm function, but I am not sure, if it is correct.
model<-lm(log(growth(Income2005,Income2017))~log(Income2005),data=Income_EU28)
ols_test_f(model)
Is there any other way to check the data for heteroskedasticity and autocorrelation if I am using betaconv.ols?
Many thanks