by Philip H. [Disclaimer:
The opinions expressed in this blog piece are the opinions of the author alone. They do not represent the opinions, policies, procedures, or sentiments of any government agency for which he currently works or has worked in the past. If you disagree, please contact the author, not your Congressmen.
]
Now, here's some interesting news about American voters. It turns out that they are not, generally, policy wonks. Nor are they truly ignorant about important issues. Rather, they vote based on a small handful of issues, often selecting party affiliation based on emotional and familial traditions. This explains a lot. It explains why Democrats are generally Democrats, Republicans are generally Republicans, and third parties (like the Greens) are not doing well in America. It explains why so many Republicans voted for George W. Bush twice, even though by the second election it was clear he was neither compassionate nor conservative (at least in the fiscal sense). And it explains why one or two issues usually define an election, no matter what qualifications or expertise the candidates bring to the plate. After all, to quote fellow ex-pat Cajun James Carville "It's the Economy, stupid!" Equally interesting to me as a scientist and framer of scientific messages, is how the writer talked about the academic work she reported on. She talks about academic conclusions from a multi-year study "couched in academic understatement." My guess is the social scientists here were doing the usual, scientifically correct thing and describing their data and conclusions within the statistically appropriate confidence intervals. Probably something along the lines of: "our results appear to apply, statistically to the American population within a 95% probability. Alternately, bootstrapped ANCOVA without regression might have yielded..." That may be correct in a talk to the National Academy of Public Administration, but somehow it always leads newspaper reporters.... to wonder what the academics area really saying, and try to get "other sides" of the story. In other words - this is bad framing for an general audience. Thankfully, in this particular case, the "fairness in reporting" stchick works - and it contributes to the reporting. In the case of Creationism/Intelligent Design vs. Evolution, it doesn't. I know, I know, scientists hate to make firm conclusions when the data do contain the possibility of error or omission. They even hate to make black and white statements when there is a really LOW probability of error. I took those courses too. But this is a public policy debate. It is about how to get American voters more engaged, or if they can be more engaged. And if the truth is Americans are one or two issue voters who inherit their political allegiances like a house or a trust fund, those facts tells us something. And no one will fault the scientists for saying so directly, and with out describing the confidence interval.
