Archives

NUMBERS JOCKS. Practically every day, some news organization or other presents us with a dire headline that includes the phrase "Study finds." Without a full exploration of the content, each of these can function as a factless propaganda item that reinforces an existing belief or annoys us by contradicting an existing belief. The interesting question is whether we, as individuals, behave differently depending on which of these two circumstances applies. Today, I came across this teaser in Drudge:


I happen to believe that this is true. It seems reasonable to think that there would be fewer accidents if all those businesspeople, housewives, and teenagers weren't motoring down the highway yakking on the phone. On the other hand, my libertarian inclinations resist the notion that the cops should have yet another excuse to punish people for what they do on their own private property (i.e., in our cars) before any external driving misconduct occurs. So I was curious to see what the "study" had found. The opening paragraphs certainly sounded authoritative:

Imaging tests show the brain directs its resources to either visual input or auditory input, but cannot fully activate both at the same time, the team at Johns Hopkins University in Baltimore found.

Johns-Hopkins. They're smart. They must know. But then Reuters briefly describes the methodology:

Writing in the Journal of Neuroscience, Yantis and colleagues said they tested people aged 19 to 35 by showing them a computer display while they wore headphones playing voices.

At the same time, the volunteers brains were scanned using functional magnetic resonance imaging.

They were told to look for specific numbers, for instance, on a computer screen, while hearing recorded voices saying a stream of numbers...

This, the article concludes, "is like driving and trying to talk on a cellphone."

Uh, no, it isn't. It may be that there is some fundamental validity to the point that driving and talking on the phone are competing tasks which the brain doesn't handle efficiently at the same time, but this test proves no such thing. It's a setup. The exercise performed by the volunteers involved identifying the exact same information, specific numbers, via two different senses, sight and hearing. The one task is a direct and highly specific confusion factor for the other. What kid hasn't teased a pal who was trying to memorize a phone number from the yellow pages by reciting a stream of random numbers out loud? In this case, it's obvious the brain will switch its number processing capacity from one input source to the other, aware that no mixed attention span can do justice to either input.

This is very far from being akin to the circumstance of driving and talking on a cellphone. Neither task is directly analogous to the processing of numbers, which occurs in a different part of the brain from language processing. If you doubt this, consider the difference between the number task -- identifying specific numbers -- and the phone conversation task, which can lose or ignore plenty of specific words and still achieve understanding of the gist. And the brain process that attends to the road is neither numeric nor linguistic -- it's performing a kind of subconscious pattern recognition, looking for important exceptions in the overall visual field which require the application of conscious decision making. If it were otherwise, we would find it impossible to do what we all do on a regular basis: drive some familiar route with hardly any recollection of the trip after we've reached our destination. No exceptions spotted, no conscious processing required.

The study focuses on one kind of processing function, not related to driving or talking, and tries to compare it to a situation that involves two altogether different kinds of processing functions which may or may not create complications in an organ well known to be capable of advanced multiprocessing tasks.

The conclusions are nonsense. Johns-Hopkins should know better, and we should all be reminded of the need to be skeptical when we're urged to be credulous of some scientific authority and even when we're inclined to agree with the conclusions being reported.

Is this how the link between global warming and human behavior came to be regarded and reported as a "fact"?