Lies and statistics

Benjamin Disraeli once said, “There are three kinds of lies: lies, damned lies and statistics.” I’m not sure which category the following falls into, but I definitely find it hard to believe. Researchers from Duke University, Baruch College and Loyola College of Maryland collaborated recently on an experiment that purports to show that restaurant chains do much more harm than good when they add healthy items to their menus. According to the researchers, many consumers make less healthy menu choices when a healthy item such as a salad is added to a menu than when the menu does not contain the healthy foods.

According to Australian Food News, where I read this account, research leader Gavan Fitzsimons, professor of marketing and psychology at Duke University’s Fuqua School of Business, calls the phenomenon “vicarious goal fulfillment.” Basically, Fitzsimons says, a person can feel they have met a goal if they have taken some small action—in this case, it would be considering ordering the salad. Having weighed the option, the person then chooses the least healthy item on the menu.

The experiment was conducted by showing research subjects one of two picture menus. Half of the participants saw a menu with only unhealthy items, while the other half saw a menu with the same items, along with a side salad. The researchers said they found that the great majority of subjects chose an unhealthy menu item over the salad, and that the choices made by people who saw the menu with the salad were actually less healthy than the choices made by the people who did not see salad on the menu.

“In this case, the presence of a salad on the menu has a liberating effect on people who value healthy choices,” AFN quoted Fitzsimons as saying. “We find that simply seeing, and perhaps briefly considering, the healthy option fulfills their need to make healthy choices, freeing the person to give in to temptation and make an unhealthy choice.”

The researchers conducted a similar test with a menu that featured a bacon cheeseburger, chicken sandwich, fish sandwich and a veggie burger; and several types of Oreos against a 100-calorie pack of Oreos. The results, they said, were essentially the same.

“What  this shows is that adding one or two healthy items to a menu is the worst thing you can do,” Fitzsimons said.

I disagree, and I believe research such as this does a disservice to the foodservice industry. Although it is quite possible that adding a healthy item to a menu may not make much of a dent in your sales of burgers and fries, it hardly proves that adding unhealthy menus is a bad thing. If I am in the mood to eat healthy, and the restaurant I choose has only one healthy item on its menu, if that item is not appealing to me I am not going to choose it. People who want to eat healthy don’t want to be treated like second-class citizens. They want to have options, just as a vegetarian doesn’t want to have only one non-meat item on the cafeteria menu.

In his report, Fitzsimons noted that although McDonald’s continues to report strong sales, they attribute their strength to the Quarter Pounder, not to healthy choices. Forgive my cynicism, but I know that in the right hands to can make statistics say whatever you want. And even if the data have not been manipulated, does that mean that McDonald’s healthy options aren’t making a difference?

My suggestion to Professor Fitzsimons is this: get a group of subjects together who say they want to eat more healthfully. Show them two menus, one with only less healthy choices and one with a variety of healthy and less healthy options. Note their choices from each menu, and see what you conclude from the data. Then let’s talk.