The classic example for the fundamental principle that correlation does not prove causation is the ice cream/crime connection. Daily spikes in the rates for aggravated assault are correlated with spikes in sales of ice cream. Therefore, ice cream causes crime and we should ban it, right? Wrong. Hot weather causes people prone to assault to commit more assaults and people who like ice cream to buy more ice cream.I agree with Kent's critique of correlation vs. causation, yet I've had some personal experiences that warn me not to dismiss quite so quickly the candy-crime conundrum.
The reason the example is classic is that the hypothesized direct causal connection (that ice cream causes crime) strikes us as so patently absurd (in jargon, lacking facial validity) that the students instantly know something is amiss. Where the causal connection is plausible, though, we see exactly the same fallacy put forth and accepted by too many too often. The correlation between poverty and crime "proves" that poverty is the root cause of crime, for example. Too many people see no need to probe further.
Now we see a serious proposal that comes close to the classic example.
Many years ago now, Kathy and I became the guardian of a young 11-year old neighbor who we'd known since early childhood after her father died and her mother went to prison. The young lady (who today is grown with a child of her own) had run away from her relatives after her parents were gone, and she'd been either a runaway on the streets or locked up in juvie hall for months before we took her in.
To call this child's behavior atrocious when she first arrived in our home understates the case; on some days the task had more in common with taming a wild animal than what you'd typically think of as parenting. She suffered tremendous grief, pain and anger, and as a long-neglected kid would lash out and act up to get attention (negative attention is often as good as positive for neglected children), creating an ugly, self-perpetuating cycle.
We did everything we could to try and break that cycle and one thing we noticed that really made a difference was diet. Having basically fended for herself since early childhood, she'd consumed sugar 24-7, from sugar cereal in the morning to sugar soda at night, and it visibly affected her behavior whenever we could successfully reduce it. Many times we noted a direct correlation - nervous, hyper, willful behavior within just a short time after some big sugar rush. It was so predictable Kathy and I would openly discuss it while it occurred. You could also tell when she'd "come down" from a sugar high.
We were never able to completely wean the child from sugar, and when she was old enough to be out and about on her own she reverted to her bad eating habits. But we always adamantly believed it made a big difference.
Cut to the present day, when my goddaughter's own daughter is now three years old, calls me "Grandpa," and is enrolled in a delightful Spanish-language immersion daycare she began just this summer. She loves her school and especially the gal in charge of her age group, but not long after she began there they began reporting minor behavior problems - nothing too out of the ordinary for a child then still in the terrible twos, but enough to where they sought parental help to reduce it.
Conversation didn't cut it and her mom was becoming frustrated, until she and Kathy hit on the idea of reducing sugar intake. This has worked like a charm. It had an effect within just a few days after eliminating most refined sugar from her diet - especially candy, sugar-filled breakfasts and soda - and judging by daily reports from the daycare, it has all but completely resolved (knock wood!) the problems they were seeing. Nobody - her mom, us, her teacher, anybody - can identify anything but the change in diet that accounts for the altered behavior.
Indeed, we've already been discussing Halloween with some dread, fearing that a night that generates a sack full of candy could generate weeks worth of misbehavior in the aftermath. Plots are already being hatched by others to come up with a low-sugar Halloween workaround.
We've talked a lot in my household about this phenomenon. Kathy theorizes that many of the kids diagnosed with ADHD or whose behavior is controlled through Ritalin or other psychoactive drugs at an early age are really just ginned up on too much sugar and actually need a better diet more than medication. I recall her expressing that view once to a relative who's a schoolteacher, who replied with an anecdote about removing sugar-candy from the vending machines at her school because of its effects on student behavior. Clearly we're not the first ones to notice this relationship.
One also wonders if there's a sugar-candy relationship to research-based findings (discussed here on Grits) that "Treatment for kids with signs of hyperactivity, when it's delivered by age three, can decrease the chance they'll land in the juvenile justice system by 75 percent." Hyperactivity and aggression at that age, according to these studies, are statistically significant risk factors for kids later getting wrapped up in the juvenile justice system. Might some of those "hyperactive" kids just be ginned up on too much sugar? I have to wonder.
So in the end, I agree with those critiquing the instant study, but at the same time I think the relationship between sugar and juvenile behavior is strong and not especially positive. Is it as big a factor as crappy parenting? Probably not. But I'm convinced that reducing sugar in a kid's diet can make a big difference in juvenile behavior at the margins.
UPDATE: I knew I should have run this piece by Kathy before hitting "Publish"! She informs me that it's not just sugar she believes is the culprit but more specifically high-fructose corn syrup and perhaps most importantly, certain widely used dyes that are common in America but have been banned in the UK. She pointed me to this article calling dyes the "Secret Shame" of the US food industry and regulators:
Kathy said it's hard to tell whether reducing high-fructose corn syrup or artificial dyes has been the biggest contributor to our grandaughter's improved behavior, mostly because they're so often in all the same foods. Like these:
Artificial dyes are particularly prevalent in the sugary cereals, candies, sodas, and snack foods pitched to kids. For instance, General Mills' Fruit Roll-ups and Fruit-by-the-Foot flavored snacks get their fruity colors from Yellow 5, Yellow 6, Red 40, and Blue 1. General Mills' Fruity Cheerios, Lucky Charms, and Trix also contain several of the problematic dyes, as do Kellogg's Froot Loops and Apple Jacks and Post's Fruity Pebbles.
More than a dozen American varieties of Kraft's Oscar Meyer Lunchables kids' meals contain artificial food dyes, but not so the British versions. Starburst Chews, Skittles, and M&M candies—all Mars products—contain the full spectrum of artificial colors in the U.S., but not in the U.K., where the company uses natural colorings. Even foods that aren't particularly brightly colored can contain dyes, including several varieties of macaroni and cheese and mashed potatoes. Betty Crocker's Au Gratin "100% Real" Potatoes are partly not real, colored as they are with Yellow 5 and Yellow 6, both derived from coal tar. Remarkably, in Britain, the color in McDonald's strawberry sauce for sundaes actually comes from strawberries; in the U.S. it comes from Red 40.
"The science shows that kids' behavior improves when these artificial colorings are removed from their diets and worsens when they’re added to the their diets," said Dr. David Schab, a psychiatrist at Columbia University Medical Center, who conducted the 2004 meta-analysis with his colleague Dr. Nhi-Ha T. Trinh. "While not all children seem to be sensitive to these chemicals, it's hard to justify their continued use in foods—especially those foods heavily marketed to young children."