This refers to the basic probability writeup at

The code used for the population sd in the Section 8(z-scores) is:

pop_sd <- sd(data$Screen_size.in.cm.)*sqrt((length(data$Screen_size.in.cm.)-1)/(length(data$Screen_size.in.cm.)))
So, this translates to pop_sd= sample_sd*sqrt((n-1) / n) = sample_sd /{sqrt(n/(n-1))}

As per the CLT theorem, the standard deviation of the sampling mean is given by (sigma)/sqrt(n). Now, of course, when calculating the sample sd, we may take (n-1) in the denominator. But the sd is already calculated in the code… so where is the need to take this formula and adjust again for (n-1)???