4 ms·
> (In layman's terms: not only is the bell-shaped curve shifting over to the right, it's also getting squashed down flatter and wider.) Maybe you've already ac
by kqr 5y ago
> (In layman's terms: not only is the bell-shaped curve shifting over to the right, it's also getting squashed down flatter and wider.)
Maybe you've already accounted for this, but I'm bringing it up anyway: increasing variance doesn't necessarily mean flatter and wider bell curve. There are at least two more options:
- A bell curve with a more pronounced peak, lower shoulders, and more mass under the tails. This would correspond to experiencing even more normal weather, but the extreme weather that does occur is insanely out of whack with normal.
- The peak separating into a low number of different peaks which drift apart. This would correspond to weather flip-flopping in its behaviour based on time or location.
I'm not saying this is what happens, I'm just saying we ought to be mindful of automatically committing the Gaussian fallacy whenever we hear summary statistics!