I'd want to see a definition of race based on genetics that formed a coherant group. Treatment of differing phenotypes tells you about different cultures, but doesn't help much with genetic groups.
I'd also want to see an IQ test that was independent of educational level generally, and a history of test-taking in particular, rather than being highly correlated with it. I'd also want to see an IQ test that didn't use measures of social conformity as a proxy for intelligence.
Finally, I'd want to see a study of racial groups that demonstrated that membership of the group was a significant predictor of an individual's intelligence.
OK, I am doing a lot of research on race and intelligence, and maybe I can help you, but I may need some clarification on a few of these points.
Let's start with this one:
"I'd want to see a definition of race based on genetics that formed a coherant group. Treatment of differing phenotypes tells you about different cultures, but doesn't help much with genetic groups."
The science of human population genetics tends to avoid the word "race" and instead use the word "ethnicity", which seems to have followed from the odd politics in academia in the last 60 years: there has been an effective campaign to deny that "race" is biological, only cultural. Nevertheless, geneticists have found significant biological correlations among "races", but, since they don't use the word, "race," they use other words like "ethnicity" to denote these populations. It has resulted in something seemingly absurd. Among the public, "race" is biological and "ethnicity" is cultural, i.e. you can be adopted into an ethnicity or you can marry into an ethnicity but your race is the race you were conceived and born with, impossible to change. The public tends to have their vocabulary straight. But, when you use those words in science, nobody knows what you are talking about. It is a step backward. Science is supposed to be clearer than the public. Just so I can be clear, when I use the word, "race," I mean something purely biological, and "ethnicity" is purely cultural. Ethnicity of course correlates with race, but, when you are looking at genes or phenotypes, then you are dealing with biology directly, not culture directly. I define "race" as a group of organisms with a tendency of common ancestry corresponding to common ancestral geography, a group that is larger than a nuclear family but smaller than a species. Races are best identified with frequencies of genetic markers. Unlike other taxons, races are fundamentally spectral, like colors on the color spectrum, identified with gene FREQUENCIES and TENDENCIES, not absolute genetic delineations. And, unlike other taxons, there can be "races" within a "race" (i.e. Ashkenazi Jewish race within the Jewish race within the Semitic race within the Caucasian race). So, maybe you can clarify your point for me. What do you mean by "race"?