Watts Up With the Internet? Motivated Bias on Climate Skeptic Blogs

Wed, 2011-07-13 08:06Chris Mooney
Chris Mooney's picture

Watts Up With the Internet? Motivated Bias on Climate Skeptic Blogs

Recently, I’ve become aware that the prominent climate science skeptic blogger Anthony Watts has been challenging a number of my posts. Maybe it’s because in my most recent book Unscientific America, I made a big deal about a site that attacks climate science, like his, winning a “Best Science Blog” award.

Anyways, Watts has gotten me back. Based upon my photo, he has taken to calling me a “kid blogger”  (see here and here). And it’s true: I’m 33, obviously too young to be fooling around on the Internet.

The attention is flattering—but I’ve also grown intrigued by what happens on Watts’ blog when he criticizes something or someone and his many commenters then follow suit. Because it does indeed show what a dangerous place the Internet is for kids like me.

 Watts commenters are an interesting bunch—in many ways, I’m very impressed with them. They are certainly highly energized to debunk climate science, and they bring a lot of intellectual abilities to the task.

At the same time, however, the debunking they conduct is overwhelmingly one-directional. By and large, these commenters are practicing “motivated skepticism” and showing a “disconfirmation bias” (see the image above, from this cool post) rather than conducting an open-ended informational search that could potentially end with their prior views and assumptions either being confirmed or disconfirmed.

As an example, let’s take Watts’ latest post, which is a response to my recent post on a study on astroturfing in the Journal of Business Ethics.

Watts suggests, in his post, that the researchers have done something unethical in their study. His headline is, “Researchers set up fake global warming websites to study response,” and in it he makes this charge:

So, they setup fake websites to gather fake data. Nice. Not only that, they “borrowed” content from other websites to use on these “fake” websites, apparently without citation or attribution, lest that taint the results. Sounds like a job for John Mashey and “Deep Climate” aka Dave Clarke. I’m sure they’ll get right on the case like they did with Wegman.

 So, this study seems perfect for a business ethics journal. Glad to see that the study of opposite views fits in to this trend recently published by Security Week.

Cybercriminals Creating 57,000 Fake Web Sites Every Week

Watts is thus accusing the researchers of something pretty serious…and soon his commenters come in and proceed to bash the astroturfing study. They post and critique the abstract and various passages, they check up on the authors and their funding sources and their universities—and they reiterate Watts’ critique, sometimes in far harsher terms:

So, the lying liars set up fake webpages to push their lies about humans destroying the environment and got busted at it. Good. Lying liars who lie about humans destroying the environment deserve to be dragged into court for stealing content from real people’s webpages for their lie pages.

Eventually, someone posts a link to an online version of the full study. Then, at 6:49 pm, one commenter who seems to have actually read it realizes that the whole thrust of the critique is wrong. But even he only notes this in passing, by way of launching yet another critique:

I was originally concerned, as apparently Bernie was, as well, that the researchers had located and manipulated naive web surfers…. but apparently they recruited students who were told they were taking part in an study. That’s the good news. The Bad news is that the students were told they were taking part in an evaluation of web site designs: in other words, they knew that all of the web sites they were looking at were in fact fakes or prototypes. The seemingly anomalous conclusion that students evaluated the “astro-turf” websites as non-credible but nonetheless accepted the information indicates to me that the students were evaluating the information separately from web site design, which they were supposed to be evaluating. [Italics added]

Observing all of this, I contacted one of the authors of the study in question, Martin Martens of Vancouver Island University. Here was his (highly predictable) response to the charges above:

The fake web sites were not on-line in a way that permitted viewing by the general public. They only existed within the computer system used for the experiment. The only people who saw the web sites and answered the survey questions were the participants recruited for the study.

The study was also approved by an ethics committee (of course) and when it was over, Martens explained, the participants were debriefed about it and “provided information to remove any mistaken beliefs that might have developed as a result of reading the web sites, and an explanation as to why the procedures were necessary for the experiment.

In short, this is very similar to many, many social science studies, including some true classics–like this paper on biased reasoning:

People who  hold  strong opinions on  complex social issues  are likely to examine relevant empirical evidence in  a  biased  manner.  They  are  apt to  accept “confirming”  evidence  at  face value  while  subjecting  “disconfirming”  evidence  to critical evaluation, and  as a  result  to  draw undue  support  for  their initial positions  from  mixed  or  random empirical  findings.  Thus, the  result  of  exposing contending  factions  in  a  social  dispute to  an  identical  body  of  relevant empirical evidence may  be  not  a  narrowing  of  disagreement  but rather an increase  in  polarization.  To  test  these assumptions  and predictions, subjects supporting  and  opposing  capital  punishment  were  exposed  to  two  purported studies, one seemingly  confirming and one  seemingly disconfirming  their existing beliefs about the deterrent  efficacy  of  the death penalty…

Yup, “purported studies.” They weren’t real. They were created for the experiment–a classic experiment that revealed how people who start out from different ideological positions will read the same “evidence” vastly differently, rating a study that seems to agree with them as convincing and a study that doesn’t seem to agree with them as unconvincing—even when both studies are made up and have the same strengths and weaknesses!

I didn’t choose this study by accident, of course–I chose it because it helps to cast some 100 watt light on what Watts and his commenters are up to.

Some particular piece of evidence—in this case, the astroturfing study—was flagged as disagreeing with them. So like good motivated skeptics, they went on the attack and started criticizing. Along the way, a few  caught on to the fact that the original criticism wasn’t even right…and kinda noticed…but they quickly moved on to new criticisms. 

Given all this, any predictions about what they will say about this post?

But hey, go easy on me…I’m just a kid, after all. 

Previous Comments

NikfromNYC is a rabid cliamate denier drone who surfs the web for articles about climate change and drops an inane pile of poppycock into mix. Lucky for us, he rarely sticks around to engage in the discussion. DeSmog Blog would do well to ban him from posting on its webiste.

Pages

[x]

For more than a year, oil giant BP has waged a massive public relations battle to convince Americans that the company has been bamboozled by the oil spill claims process relating to the 2010 Deepwater Horizon oil rig blowout.

This BP PR campaign has involved full-page newspaper ads paid for...

read more