Well, at least Anthony Watts apologizes for calling me a kid--kind of:
I will apologize to Chris Mooney though for calling him a “kid blogger” based on that youthful photo he uses. It just seemed so much more cuddly (he looks amiable and likable in it) than calling him a schill blogger.
That's verging on kind, until the end. The photo is from last year. Anyway, Watts does not apologize for suggesting that the authors of the astroturfing study (which is getting a heck of a lot of free PR!) committed some sort of ethical breach in their research design, which they didn't. Instead, he's off on a new critique of the study (I told you it would be disconfirmation bias without end), which he has read this time. "Why wouldn’t they mention that the study was conducted on a private Intranet and not on the World Wide Web?" Watts asks. Maybe because it's obvious? Or at least it is to me, in that I know something about this kind of research, and I read the paper, and never even got the remote impression that this was a live study on the real live web. I mean, how would you control such a thing? How would you know who your subjects were? How would you know how the websites affected them? Etc etc etc. But Watts continues:
Well I don’t know about you, but if you want to learn about something in the wild, you generally study it the in wild. What we have here are manufactured, “fake” websites, running on an Intranet (apparently, according to Mooney’s query of the authors). And generally, when I hear about a study on websites as applied to real websites viewed on the world wide web, I expect the study would be about real world websites, not one limited to a lab fishbowl. As I see it, this would be like doing Jane Goodall like studies of wild chimpanzees based on chimp-robots made to look like chimpanzees, confined in the lab, and studying how they interact with students who are told they aren’t actual chimpanzees, but disguised as marketing salesmen.
Can you follow that? I'm having trouble, but then, I'm just a kid. Let me just make the obvious point that, in order to do a social scientific study, the design has to be practical and manageable enough for you to be able to get enough participants, and learn enough about them, to actually find some strong results. This, of course, is why so many of these studies are done on undergraduates who volunteer (or get paid some small amount) and come into a campus lab. The design surely has limitations (what design doesn't?) but it is a time honored one. In any case, here's the bottom line on the astroturfing study (discussed in more detail here). Astroturfing is real, it affects the climate issue, and astroturf websites and organizations make scientific claims about climate that are misleading. They're not the only ones doing so--nobody has suggested that Watts' blog is an astroturf site, for instance--and those involved probably believe they're good and honest people who are just following the evidence, in the same way that almost everybody thinks of themselves. But nevertheless, what this study shows is that for most people, if you visit a website that contains these sorts of dubious claims, it misleads you. It makes you doubt. Perhaps this was obvious anyway, but the study captures it. Now, I think you would probably find the same thing if you didn't call it "astroturf" and just called it "climate skeptic websites" or "climate denial websites." But the point remains: Misinformation works. Not a surprising thing--but definitely a depressing one.
