Can someone explain its significance and maybe a simple example as well?
In: 3585
Let’s say that you want to know whether Steph Curry is a better shooter than Shaq. Curry makes 3pt shots at a better rate than Shaq, and (let’s say) Curry also makes layups at a better rate than Shaq.
The paradox is that while Curry is a better shooter than Shaq in both categories, Shaq has a better combined shooting rate than Curry. The explanation is because Shaq takes way more layups than 3pt shots, and layups overall are higher percentage than 3’s.
In other words, Simpson’s paradox is when you’re measuring something that looks better in both Group A and Group B individually, but looks worse in when combined. It happens because there’s more of one group than another when comparing across treatments.
So it’s an example of confounding variables where you need to identify the groups that are secretly influencing your comparisons between treatment and control.
Latest Answers