Let's make up an imaginary field of study and call it boulderolology.

Imagine it consists of scientists trying to explain the acceleration with which boulders roll down hills.

Different boulders, of course, are measured as rolling with different accelerations, and suspend your disbelief.

Imagine everyone really wants to know why.

Some of the scientists look at the size of the boulders.

Some of them look at the local temperature or air pressure.

Some of them look at the ways boulders are formed.

Others distract themselves with the history of boulders, which doesn't really help the quest much,

but is still quite interesting as a means of hypothesis generation.

Some of them look at how round the boulders are. Some look at humidity. Others look at altitude.

And these boulderolologists, they find correlations, they take their low P values, and they give

TED talks and get book deals and give advice and make generalizations, and though sometimes for individual boulders, their findings matter a lot.

Overall, none of their subfields and pet theories are able to consistently explain that much of the variance in boulder acceleration.

This is kind of how a lot of psychology feels right now.

We have scientists trying to explain differences in human behavior with all sorts of

variables, right? Are they a man or a woman? Did they have a stay-at-home mom? Did they go to a

private school or a public school or a charter school? Where they homeschooled? Were they

breastfed? Did they play violent video games school or a public school or a charter school where they homeschooled? Were they breastfed?

Did they play violent video games?

Does their area have high income inequality?

Did they have two dads?

What type of diet did their parents raise them on?

...