April 26, 2004

Chewing gum and memory

No, this isn't an American Proust post -- it's about a headline story two years back, when it was claimed that memory is improved by gum chewing. I missed the story at the time, but a friend recently mentioned that it had inspired her daughter's kindergarten teacher to hand out gum to her students -- which naturally spurred me to take a closer look.

As it turns out, the study behind all the subsequent reports involved all of 75 research subjects, all adults, divided up into three groups. One group chewed gum; another was instructed to simulate chewing; another did nothing. After three minutes, members of all three groups took a 25-minute series of tests.

I haven't been able to consult the actual published study, but from the news reports it sounds as if there are many potential problems with the study design. Perhaps most importantly, the study was intended to prove a hypothesis, so testers' bias (conscious or not) can be taken as a given -- yet there is no mention of any measures taken to neutralize such bias. Then there is the well-known effect of researcher interest, where special attention given to test subjects is itself sufficient to raise their performance. To this I would add the possibility that the physical and mental exertion resulting from being assigned a task, however minor, might well temporarily sharpen mental faculties, while being left to sit doing nothing might have the opposite effect. Finally, the study size is awfully small for its findings to have any statistical significance -- and I can find no mention of any studies corroborating the findings with larger samples and corrected methodologies.

I did find, however, another sort of followup carried out as a 2002 course project for USC's Biological Studies 230. Subjects were divided into just two groups, chewers and nonchewers; group size was just over twice as large as the original UK study, and testers appeared, if anything, to be skeptical of the original study's hypothesis. No difference in test results between chewers and nonchewers was found.

Posted by David on April 26, 2004 2:32 PM

Comments
Post a comment




  Remember Me?


(For bold text to display correctly, please use <strong>, not <b>)




Google