Markus Strasser

Beware of Proxy Chains

Often there is a proxy chain to get to the quantity of interest

"Happiness is 50% determined by genes"

flaw: terms themselves are not well defined. Stated and unstated feeling; language compressor; personal thermometer; siamese twin example; state of data capture; needs an explanatory theory of qualia

If genes code for thick butts and that makes girls happier than what if society's beauty definition changes -- all your science is out the window

Usually you get the people equivocate: "but the effect is still there", but that's just talking about correlate and not causal. It’s not an explanation

"Happiness is 50% determined by genes"

would explain something if genes can link to objective properties that people point to when they talk about happiness

If an experiment is confounded (poor experiment design), if it produces a strong signal even when its experimental hypothesis is true, then using a larger sample size will just make that signal even stronger.

SSC bad experiment designs

I run studies on things like number of arms, number of kidneys, size of the pancreas, caliber of the aorta, whether the brain is in the head or the chest, et cetera. 90% of these come back identical – in fact, the only ones that don’t are a few outliers like “breast size” or “number of penises”. I conclude that men and women are mostly physically similar.

In programming, you can catch some errors with a type system but the most important errors are the ones your type check and tests can't catch

Beware of Proxy Chains