What is the difference between computing manually some parameters from my model summary and doing the same thing through hypothesis testing? Since sigma’s in a Normal single-level model are in log-scale, I have to run: exp(sigma_Intercept) and exp(sigma_Intercept+sigma_treatment1) in order to compute the residual standard deviations of my two treatment groups. However, when acting similarly with the Hypothesis testing, I get slightly different values. The hypothesis is the following: c(“exp(sigma_Intercept) =0”," exp(sigma_Intercept+sigma_treatment1)=0" ).
Thank you in advance.