Sunday, August 17, 2008

What Do You Know, You Can Get People To Agree With Themselves by Anthony McCarthy

Or, Do these guys ever listen to themselves?

In this mornings Boston Globe in yet another psych-ridden look at a story in the news, the “Clark Rockefeller” riddle, is this piece of vital information:

These subtle cues, Pentland emphasizes, are difficult to fake - he calls them "honest signals." But Bailenson and Yee's work suggests that the cues don't have to be subtle to work: most of their subjects didn't notice that they were being mimicked, even as they were proceeding to bond with the digital figure on the screen in front of them.

To earn someone's trust, in other words, even rather blatant aping can do the trick. One of the landmark studies on influence was done in 1965 by the Ohio State psychologist Timothy Brock. In it, shoppers at a paint store were approached by a research assistant who offered them advice on what type of paint to choose. He told half of the shoppers he approached that he had recently bought the same amount of paint that they were looking to buy, he told the other half he had bought a different amount.

By and large, the first group took his advice, and the second did not. Something as trivial as buying the same-sized bucket of paint, Brock argued, can forge a bond with a total stranger.

I had to read this at least four times to believe I was really reading this in the masthead article in a prestigious section of a major newspaper. I thought it was incipient dementia at first.

Apparently it takes a “landmark study” for psychology to understand that if a store clerk tells you he did what you wanted to do you’re likely to do what you wanted to do. And if he tells you he didn’t do what you wanted to do, you’re likely to do what you wanted to do. In short, people tend to do what they want to do.

Apparently this piece of obviously counterintuitive wisdom is impressive enough to be used in the Boston Globe Ideas Section forty-three years after it was published.

I’m tempted to look it up the ancient study to check to look if they controlled for:

- people who tried to estimate how much paint they would need and the methods they used,

- how many of the buyers just went into the store and guessed,

- whether or not they were trying to cover a darker color with a lighter color,

you know, the kinds of things real people think about when they’re painting something. In the small samples typical of studies of this type, even a few people taking these kinds of things into account might give entirely different results. But I’m on a very tight schedule. If you know the study, please tell us.

I’ve got a hunch that too much dependence on psychology is a sure sign of a lazy reporter, the Boston Globe Ideas section would be a good place to look for evidence.

II. No Joke

You might also read the letter by the executive director of public and member communications of the American Psychological Association, denying that it has done little to keep its members from actively participating in the Bush regime’s use of torture. Contrast it with this news article by Tania deLuzuriaga in another place in today’s Globe.

This issue, brought to new prominence by Jane Mayer and especially her reporting on the role of Martin Seligman in relation to the torture scandal. It’s tempting to be sarcastic about his protestations that his lecture to the incipient torture team was turned on its head and that his torturing dogs was for the greater good but that might risk turning this into a joke. It’s no joke.

What did Seligman think they were asking him for? I don’t believe he couldn’t have guessed that his lecture and the dog torturing experiment was very likely to be used to plan a program of torture in the year after 9-11. If he’s that naive you wonder what his lifetime in psychology has taught him about the way people think. A highschool history student looking at the darker side of the 20th century for a term paper would likely have guessed better. But then, I think history has a better track record for getting that kind of thing right, than psychology has. I also don’t trust people who torture animals for the greater good.


And finally, a very rare and skeptical look at the use of imaging in studying behavior by Jonah Lehrer.

"There are so many bad brain imaging studies, it's hard to believe," says Nikos K. Logothetis, director of the Max Planck Institute for Biological Cybernetics in Germany. "Too many of these experiments are being done by people who, unfortunately, don't really understand what the technology can and cannot do."

Logothetis and others believe that much of the misuse stems from the visual nature of the data. One study, by researchers at Colorado State University, showed that simply giving neuroscience students images from an fMRI machine, even if the images were redundant or irrelevant, made the students significantly more likely to find the data credible. According to Paul Bloom, a psychologist at Yale, this is because fMRI "has all the trappings of work with great lab-cred: big, expensive, and potentially dangerous machines, hospitals and medical centers, and a lot of people in white coats."

The data looks rigorous - it has the veneer of cutting-edge science - and people assume it's valid, even when the reasoning is shoddy.

"You can't just put people in a scanner and ask them whatever question you want," Logothetis says. "Many of these [fMRI] papers are such oversimplifications of what's happening in the brain as to be worthless."


n e-mail objects most strenuously by to my jibe about the vintage psych study above.

First, what is there about being deceived by the accused con-man “Clark Rockefeller” and the experience of being psychologically manipulated by a psych student posing as a store clerk in an Ohio paint store in 1965 which makes the latter intuitively relevant to the former? What does the study of 1965 authoritatively add to the intellectual investigation of the bizarre news story? Please forgive me if I assert that the more rational assumption is that the study has nothing to tell us about it.

The two attempts at deception are insufficiently similar, to start with. The possible motives would seem to be entirely unlike. The people attempting the manipulation are hardly alike and the subjects almost certainly as unlike, though that would be an entirely undefined variable in the comparison. The fact is that “Rockefeller” is suspected of a decades long, geographically vast and complex range of deceptions, perhaps, including killing. The “clerks” in the paint store were certainly not doing it for the same reasons. They probably just wanted to get their degree.

We know the subjects in the study wanted to buy some paint, we have no idea of what range of motives the people who fell for “Rockefeller” under his various names and guises may have had or if those deceived by him shared anything else in common. There is no reasonable reason to suspect that the situations have any real life similarities that would make resulting claims linking them to have sufficient reliability to be considered science or journalism.

About the paint store study, let me pose just two possible scenarios to you.

You are a woman who has painted a heck of a lot of things. You are a very experienced painter who has taken the time and made the effort to come up with what experience leads you to believe is a good estimate of what you need to buy. You go into a store and see a younger man (somehow I’ve got the feeling that all the psych students were probably males, though I don’t know that*) who says that he bought a similar amount of paint. He proposes a brand of paint you have had good results with in the past, so you buy it. It might have been one you were considering buying anyway.

Another scenario. Assume you are an experienced painter who has made an estimate and the kid says he bought twice as much paint. You might be on your guard for being sold too much paint, maybe someone did that to you in the past. You know the kind of paint he tells you is not what you need. You might think the kid doesn’t know what he’s talking about. You don’t take his “advice”. If he tells you he bought less than you know you’ll need, you would also think he doesn’t know what he’s talking about. Maybe you decide to shop for paint elsewhere.

The ages of the customers and the “clerks” would have just about certainly been an issue for many of the customers. At least the ages could be objectively verified, many of the other issues couldn’t.

The people who walked into that paint store might have represented any range and combination of variations on the two scenarios proposed above, and there could have been other issues relevant to the choice that was made. Maybe the kid looked like a favorite nephew or one you think is a snot nosed brat. I wonder if there is any reason to believe that the customers were all sufficiently alike to assume they would have belonged in the set conceived of in the study other than the convenience of the researcher.

Would the motivations of two people either buying (or not buying) what the clerk recommended be the same action, representing the same motivation? I doubt it. I believe their motives would have been vastly different, perhaps no two truly alike.

How large a sample would you need to take all the possible relevant issues into account? What kinds of equations would have covered all these possible issues? How would they tabulate the results?

If these are problems of sufficient complexity so as to be a insurmountable problem for psychological researchers that might be unfortunate. It doesn’t make them irrelevant. Not if discovering reality is the goal. As seen in the rest of my morning post, decades old studies can have horrible real life impacts.

People working in journalism who make the improbable assumption that these “studies” are relevant to their reporting of the news are guilty of multiple journalistic sins. Just the facts. The verified facts, supported by two independent sources. Anything else isn’t news.

* You wonder what the possible effect of having a female “clerk” advising a customer on paint would have had on the choices made in 1965. I don’t know about Ohio but do suspect that it would have been very unusual in most places in 1965 to have a female clerk in most hardware departments. You wonder at the difference in gender between “clerk” and customer. And you wonder about the various responses to that among customers of the same gender. How can you account for those contributors to the ultimate choice. And you wonder what the relevant social changes would do to the applicability of the “findings” four decades later.

I can’t tell you, just that I think they’re probably important.