Quotes of the Day

If you're looking to lose weight, here's a simple tip: don't dine with the skinny dude who stuffs his face. According to a study that will appear in the April 2010 issue of the Journal of Consumer Research, both the size and the consumption habits of our eating companions can influence our food intake. And contrary to existing research that says you should steer clear of eating with heavier people who order large portions, it's the beanpoles with the big appetites you really need to avoid. "They're big trouble," says Gavan Fitzsimons, a marketing professor at Duke's Fuqua School of Business and one of the study's co-authors.

To test the effect of social influence on eating habits, researchers conducted two experiments. In the first, 95 undergraduate women were individually invited to a lab ostensibly to participate in a study about movie viewership. Before the film began, each woman was asked to help herself to a snack of either M&M's or granola. Another "participant," who was actually an actor hired by the research team, grabbed her food first, in full view of the subjects at the snack line. In her natural state, the phony participant weighed 105 lb. and wore a size 0. But in about half the cases, she wore a prosthetic designed by an Academy Award–winning costume studio. The fat suit increased her weight to 180 lb. and puffed her clothes to a size 16.

Both the fat and the skinny versions of the actor scooped five tablespoons of food (approximately 71 g of granola or 108 g of M&M's) onto a plate. That's a heap. The subjects followed suit, taking more food than they normally would have had they eaten alone. However, the subjects took significantly higher portions when the actor was thin. During the movie — a five-minute clip from the Will Smith film I, Robot — they also ate significantly more if the actor was skinny. "It's our intuition sometimes that you don't want to eat with big people because you're afraid you'll eat more," says Fitzsimons. "In fact, the opposite is true."

What happens when a thin person takes a small portion? Again, we tend to mimic those around us. For the second test, in one scenario the actor took two pieces of small candy from a set of snack bowls. In the other scenario, she took 30 pieces. Under the lots-of-food condition, the results mimicked the first test: subjects grabbed and ate significantly more candy when the actor was thin. Under the little-food condition, the subjects took the lead of the actor and restrained their candy consumption. However, in this scenario it was the obese lunch date who posed a threat: the subjects ate more if the actor was wearing a fat suit.

Each of these tests illustrates the psychological trait known as anchoring. Humans tend to latch on to one specific piece of information when making decisions, in this case the habits of the actor. The social environment is extremely influential. If this fellow study subject is going to take an above-average number of M&M's, so will I. Call it the "I'll have what she's having" effect.

However, we adjust the influence of the social environment on the basis of how we perceive the people around us. So if an obese person is helping himself to a large portion, I'll hold back a bit because, well, I see the ultimate results of his eating habits and don't want the stigma associated with being overweight. But if the thin person eats a lot, why shouldn't I follow suit? If she can gorge herself and still keep trim, why can't I?

At the same time, if a thin dining companion orders a small portion, I too will hold back because I want to mirror the habits of a body type to which many people aspire. However, if an overweight person orders light, I'll make an adjustment. Obviously, small portions aren't working for him. If tiny meals don't help you stay trim, what's the point? Get me the cheeseburger deluxe.

• Sean Gregory
• A new study shows how the size and behavior of our dining companions can affect our own consumption habits