Science Advocacy by Polls Fails Again; Science Misleads (and Misreads) With Stats

UPDATE – 3/20

ScienceInsider has corrected the error in their post, now noting that seven percent of respondents, rather than 23 percent, chose India as the country they thought would lead in science and technology in 2020.  The rest of my post still stands.


ScienceInsider engages in a bit of scaremongering earlier today with this piece, using the tried and true (and usually ineffective) method of trotting out a poll to plead for science.  The post title is “Poll: Many U.S. Votes Have Gloomy View of America’s Science Future.” The poll was released by Research!America, an advocacy organization for biomedical research.  They have failed to adapt to the end of the doubling of the National Institutes of Health budget, so I’m not surprised they are grasping for any rationale in a storm.

The ScienceInsider piece relates the sleights-of-hand as follows:

nearly 60% believe that a country other than the United States will lead the world in science and technology by that time [2030].”

“Only 42% said they thought the United States would retain its position as the world leader in science and technology by 2020, while 26% predicted China would assume that mantle, and 23% chose India.”

Okay, so it seems easy to assume that nearly 60 percent polled (1005 likely voters) believed a country other than the U.S. will lead the world blah, blah, blah.  But go to slide four in the poll and take a look.  Twenty-three percent did not actually choose India, but weren’t sure who would lead.  Only seven percent picked India.  While the Research!America tagline (“Over half doubt U.S. leadership in science and technology in 2020”) is more accurate, the poll responses aren’t nearly as insistent.  In any case, the country that most expected to lead in science and technology in the year 2020 was….the United States.  So, why the gloom, exactly?  Because it was a plurality who thought so rather than a majority?  Somehow I think both Research!America and ScienceInsider would be rending their garments over the poll results even if it was 51 percent who thought the U.S. would lead rather than 42 percent.

Why does global leadership matter?  Probably because other poll questions indicate people think world leadership leads to economic competitiveness.  Go big or go home?  It might be more accurate that striving to be a world leader helps economic competitiveness, but if what determines the world leader in science and technology is something not directly linked to economic performance, I’m not as persuaded to make the logical leap.  For instance, if a country has 4 percent of its GDP committed to research and development, it would lead the world.  But if it’s GDP is just a billion dollars, how valuable is that number one ranking?

Even if one thinks polls on niche issues are a good way to influence public policy, over-interpreting them doesn’t do very well for the cause, or the advocate.  Shame on Science for the sloppy journalism, and for buying into a race for number one mentality that may not be the most effective use of science and technology in the national interest.  I’ll cut Research!America some slack, as it’s more explicitly an advocacy group than Science‘s publisher, AAAS.  But I’d suggest to it’s contributors that polls like this aren’t the best value for money.