If girls have never been caned in the same number as boys how can you draw on the overall effectiveness, if you are caning less girls then there could be larger number of girls it could be effective on
Because we know how to do mathematics.
If 10,000 boys are caned and 7000 have improved behaviour while 100 suffer side effects, we can say that there’s a 70% chance of positive results and a 1% chance of negative.
If 500 girls are caned, and 100 have improved behaviour while 100 suffer side effects, we can say that there’s a 20% chance of positive results and a 20% chance of side effects.
The reliability of the second set of data would be somewhat less than that of the first – but even allowing for sampling error, you can draw useful comparisons.