Comment by sshrin

Thanks for this explanation. I understood most of it but could you explain why you should normalize using 1/sqrt(n) and why doing so makes the result converge in distribution?