A new article in the 'early edition' of PNAS suggests that IQ tests are not so simply interpretable as some might believe, or wish. Duckworth et al. ('Role of test mutation in intelligence testing') show that not just the raw DNA in a person, or his/her environmental circumstances such as diet and books in the house, but also motivation affect both the score and the usefulness of the test in predicting life success.
The authors studied 2008 individuals and found substantial increases in IQ test scores by highly motivated participants, greater for those with lower baseline scores. IQ scores predicted many aspects of life outcomes, but basically not after motivation in taking the test was taken into account.
No matter how you may feel about the Nature-Nurture debate in regard to intelligence, this paper is yet another example of the complexities of biological causation. In itself, it's probably a minor study relative to the larger question of how much of what an organism is, is 'inherited' in its genome, and how much is due to many other factors of its life-experience. The authors themselves temper their interpretation.
However, the relevance to societal issues is great, since much in your life depends on how well you fit into the system that assigns resources and rewards, and because discrimination, positive and negative, results. This affects your social 'fitness' and one can debate if it affects your evolutionary fitness. But the question goes far beyond the political.
If organisms can be predicted from their genomes, then environments don't count much. That is relevant to GWAS and personalized medicine, of course, but also to evolution itself. If genomes are predictive we can effectively forestall disease. That would be good (though it's largely not true).
But what about evolution? The idea of evolution in the adaptive Darwinian sense is that genomes are predictive in the context of the environment! That is what adaptation is all about--adaptation to the environment. It is those genomes that, in their environmental context, proliferate the most.
This at its very core means that genomes cannot in themselves be predictive. Or, put another way, adaptive Darwinian evolution can only occur to the extent that genomes are not predictive of the organism's traits. In this sense, genomes evolve as respondents, not determinants. In a static environment genomes might become predictive, or seem predictive, but that would be because of the environment, not in spite of it.
The politics of IQ testing and its many, sometimes evil, heads is not our point here. Again, we just take this story as an illustration of the elusiveness of biological causation, and the subtlety of the Nature-Nurture phenomenon.