There is an update at the bottom of this post, where it will make more sense than it would at the top here.
When Ria Misra posted a couple weeks ago that more than 1/3 of Americans believed that astrology was a science, my first thought was "surely not!" Certainly the question itself must have been confusing, a sentiment shared repeatedly in the comments below.
Well, we commenters weren't the only ones who had some faith in the American people. Richard N. Landers, on his NeoAcademic blog, recently shared the results of his own little survey—which first asked people to define "astrology", and then asked if they thought astrology was scientific.
Landers' results were pretty much what we all suspected: that only those people who confuse "astrology" with "astronomy" thought that "astrology" was very scientific. Of those who correctly defined the term "astrology", only about 13% thought it was "somewhat" or "very" scientific . . . a far cry from the 42% found by the National Science Foundation.
His numbers track closely with those found by a 2005 European Commission study, which asked half its participants "is astrology scientific" and half "are horoscopes scientific". Forty-one percent of those asked about "astrology" thought it was scientific, versus 13% who were asked about "horoscopes".
So what can we learn from this? At the very least, that the National Science Foundation is bad at conducting rigorous surveys. If you want to know what people think about X, you can't always ask "what do you think about X" . . . sometimes to get the real answer, you have to ask about Y or Z instead.
(And perhaps that it wouldn't hurt for the media to show a little more skepticism about "studies" that find surprising and counterintuitive results.)
UPDATE: A second micro-study has challenged the conclusions of the first micro-study. It claims that, when asked the NSF question in the context it was originally asked, very few people would actually confuse astrology and astronomy. (This was also a point brought up in the comments under Ria's article, by people who thought the NSF results might be valid.) A third micro-study is apparently being performed (by somebody involved in the above-mentioned European study) to specifically address the differences between the first micro-study and this new second micro-study, but this third micro-study isn't done yet. Also apparently somebody created a Kinja account just to reply to me! That's kinda flattering, even if it was just to tell me I'm wrong.