Exercise

Sample size for critical region

Using the randomization distributions with the small and big datasets, calculate different cutoffs for significance. Remember, you are most interested in a large positive difference in promotion rates, so you are calculating the upper quantiles of 0.90, 0.95, and 0.99.

A function for calculating these quantiles, calc_upper_quantiles() is sown in the script.

Instructions

100 XP

As a reference point, run the call to calc_upper_quantiles() to calculate the relevant quantiles associated with the original dataset of 1000 permuted differences, disc_perm.
Do the same for the small dataset, disc_perm_small …
and for the big dataset, disc_perm_big.