in question 5, it says "Find a bootstrap confidence interval of the median of T". We were working in the notebook with sqrT so my doubt is if that's an error or we are supposed to calculate it for T instead of sqrT. In case is not an error, do we have to suppose that sigma hat is equal to the mean of the sample, so we can use the second case we saw in class?
there is no error, we were working on sqrT in order to use a Gaussian model (hence the name transformation method for part 1) but at the end of part 1, we come back to T because sqrT is of no interest really, what we want to study is T, the incubation time.
Then the second part is about Bootstrap method and one of the good points of this method is that we don't need a model (Gaussian or otherwise) and hence we can compute an IC with data from T directly.