Giving a nonsense test to people and recording a correlation between the nonsense score and genotype is science, even if the questions are bunk. Never once have I claimed that the actual test results mean anything in particular -- just that the measured score is lower on average in those with genes that express as black skin which is a scientific observation.

No, that is at best surveying. It is as scientific as asking people their favorite color or who they will vote for in the next election.

Science is working with a theory and gathering evidence for that theory. The theory behind IQ is called general intelligence and it has been completely exposed as nothing but racist pseudo-science. Removing the theory from the metrics just gives you a bunch of data, which may or may not describe a useful model of nature (but most likely it won’t).

It isn't unscientific to hypothesize that if you give people a certain set of bunk questions that based on some arbitrary scoring system that the differences will appear correlated to genotype, and then test it and find out genotype results in some weak correlation of the results. In my case, the bunk questions I am referring to are those in one or two tests commonly referred to as 'IQ tests.' It might not be scientific to decide it measures intelligence, but it is scientific to test and observe the difference.

What you're attempting to do here is debase your argument to semantics, by excluding what you don't like out of science, which is quite typical on HN when you have nada for an argument. I simply unequivocally reject this sort of 'heads I win, tails you lose' redefinition of science.

What is your hypotheses? What are you trying to predict? If you don’t have any hypothesis, if the only thing you are predicting is what you measure, you are by definition not doing science.

Science is more then mere observations. You can observe all the white swans you like, and if your only prediction is “all the swans I have seen are white” that isn’t science, it is just bird watching.

If you want to measure IQ and you observe differences in population, and you don’t give a theory as to why that is, and you don’t make any further predictins based on your observations, you are just surveying.

Now, somebody might create a model based on these observations, and make predictions based on that model, then-and-only then are you doing science. If Tycho Brave had only been observing the stars, never developing the Tychonic system, he would have just been an excellent astrologer, and as you know astrology is not a science. When Kepler later used that data to prove Kepler’s laws of planetary motion, then—and only then—would we consider his observation to be scientific.

In IQ testing, the only people making predictions with the data are psychometricians. They are using it to try to prove the existence of the g-factor. This effort is in vein, as this theory has long been debunked. It is modern day alchemy (or phrenology if you will). Nobody else is making any other predictions with IQ data. The only people who are using IQ data are using it to make predictions in a futile effort to prove their race pseudo-science. And they keep failing at it.