Rock Walls and Confirmation Hearings

Lets say you have a pile of rocks. They are of equal density and strength, on average, but are of different colors: red, yellow, blue. If you are building a wall, and want to balance colors, and you notice that the wall disproportionately favors blue and yellow, you could easily say: I need more red rock. I’ll select out red rocks, and of those rocks, find one of the right shape and density.

If those rocks are equal, and the unequal distribution has been simply a matter of a statistical “blip” then if you sort by size and density, EVENTUALLY the colors on the wall will balance out.

If you sort by “red” first, and then size and density, all is well, as the average distribution of these qualities suggests that the same proportion of red rocks will be appropriate — but you are correcting the color imbalance. More rapidly.

If, however, the average red rock is smaller or less dense, then the proper approach would be to FIRST sort for size and density, and then from that group find one that is red.

Now…all of this has been based on the notion that the original error was simply a matter of random chance going a little wonky. But what if those rocks have been selected deliberately? IF there is an imbalance, due to selection bias, then either there was neutral judgement concerning color, but unfortunately the red rocks simply weren’t dense or large enough to have much presence: or someone deliberately avoided selecting them. Perhaps they BELIEVED that red rocks weren’t dense…