The Science of Thought
Powers, Mike, Human Ecology Forum
For years psychologists have considered most children and adults to be pretty mediocre scientists. Professor Barbara Koslowski says that a broader understanding of what constitutes scientific thinking shows that we're really quite good at it.
Here's a little exercise to test your ability to think scientifically: You run into a few reports that a people who own red cars get great gas mileage. Coincidence?
Then you learn that the color red revs people up and makes them more alert. Now you have information that could explain why car color is related to fuel efficiency: red cars make people more alert, alert people drive better, and therefore fuel is conserved. Would you then test your explanation, perhaps by looking at fuel efficiency in other situations?
According to the research findings of Barbara Koslowski, an associate professor in the Department of Human Development and Family Studies, there's a good chance you would. Because, just like scientists, when nonscientists notice connections between things, they look for theories to explain why. Conversely, when they have information suggesting that two things should go together, they'll look for instances in which they do.
"We think that is pretty smart scientific behavior," she says.
That's quite a compliment, because psychologists generally consider most of us laypersons to be pretty lousy scientists.
Kowlowski begs to differ. She has just written book. Theory and Evidence: The Development of Scientific Reasoning (M.I.T. Press, 1996), that goes nose to nose with this observation. The problem, she says, is that psychologists have historically defined scientific reasoning as consisting of a very narrow set of behaviors.
"The question of whether people are bad scientists depends on whether you focus on aspects of scientific thinking that studies often ignore," she says. "There's a strong tradition in the United States and England of adopting an approach to scientific thinking that stresses only the work of the philosopher David Hume and his notions of covariation."
Covariation is another name for the correlation between two phenomena. For example water freezes when its temperature drops below 32 degrees Fahrenheit. It boils when its temperature rises above 212 degrees Fahrenheit. So the behavior and temperature of water covary.
"Simply put, covariation means that if X causes Y, then Y is present when X is present and Y is absent when X is absent," she says.
But, Koslowski argues, covariation alone is not adequate to conduct sound scientific thinking. Equally important is "theory," the investigation of different possible reasons for a phenomenon. It's a factor that psychology has tended to ignore. But without theory, she says, people would never be able to distinguish between covarying relationships that are worth investigating and those that aren't.
As an example, she jokingly uses a correlation that might be made per capita ice cream consumption and the level of violent crime.
"Even though these covary, you're certainly not going to suggest that to decrease the level of violent crime, all we have to do is prohibit ice cream consumption," she says. "That's because there's no theory or mechanism that could explain the process by which ice cream consumption could cause crime."
But suppose someone suggested a connection between fat intake and the body's production of testosterone, she asks. That theory might be enough to warrant taking a look at the relationship between ice cream and violence because it would provide such a mechanism.
"At the Centers for Disease Control, for instance, this is exactly what practicing scientists do," she says. "The first time they learned that a victim of Legionnaire's Disease was in a room that had an air conditioner, it was just an interesting bit of data. But when they learned that many victims were in rooms with air conditioners, they looked further. …