More details on the craptacularity of that study:

Hi Jonah, I’m a real-life neuroscientist from large university in Michigan and I liked the article you wrote on the recent study on why conservatives are cognitive defectives. I just read the original article published in Nature Neuroscience and I can expand a bit on some of the details. First of all, the main finding in the study is a correlation between political ideology and brain electrical signals. I think that most people are more interested in the accuracy differences between liberals and conservatives. The authors are somewhat circumspect reporting on the actual accuracy differences between liberals and conservatives. They state simply that “stronger liberalism was correlated with greater accuracy on No-Go trials (r(41) = 0.30, P < 0.05).” This is a critical finding in the paper and it is also much weaker than the brain findings. To translate the statistics, it indicates that 9% of the variance in accuracy was accounted for by political orientation or conversely that 91% of the variance in accuracy was attributable to factors other than political orientation. Moreover, they do not present a figure of this data as they did for their other results. When a correlation is this small (.30), it can be heavily influenced by the performance of a very small number of subjects who may be outliers. A related point on the possibility of a small number of subjects influencing the data. They don’t report the numbers of subjects who were liberal versus conservatives but perusal of one of their figures indicates that they had a grand total of 7 subjects who were on the conservative side of ‘0′. This is totally inadequate for any behavioral study. They also don’t report the gender breakdown of the subjects by political orientation. There is strong bias to females in the study (63%) but, as you know, there is a very strong likelihood that the males were over-represented among the conservative participants (this is not reported). This may sound like a trivial factor but it is not. I can tell you from my own experiences in testing university students in dozens of studies on cognitive abilities that the typical university male could not give a crap about how they perform in psychology lab experiments. This is a critical factor if there is a tendency for the conservatives in their study to also be males. You hit the nail on the head in your article that the largest problem with studies like this is the inferences that people draw from them. In particular the comparison between George Bush’s single minded commitment to the Iraq war with John Kerry’s nuanced openness to new information is, as you might say, nonsense on stilts. The task in the study was deliberately designed to be unbelievably boring and, hence, to bias people to errors due to inattention or “lack of monitoring.” The task involved making button-press responses to M’s and W’s on a computer screen for 500 trials. This is not a typo…500 trials. To take this kind of mind numbing task and imply that it has anything at all to do with consciously reasoning and evaluating massive amounts of information about a war, doing some kind of cost-benefit analysis in consultation with dozens of advisors over periods of weeks or months is foolish. A closer real-life application of this kind of task might be an air traffic controller who examines the same blips on a screen for hours on end and still has to be able to pick out an odd-ball occurrence on the fly. In fairness to the authors of the paper, they do not make any fanciful interpretations like this (at least not in the journal article). It was other researchers who were not involved in the study who were drawing the silly inferences that were published in the press account on which you based your article. On the other hand, it is still important to address these faulty inferences as this is the interpretation that your average journalist or blogger is going to make.