Peer modeling effects in the smoking behavior of teenagers.

Modeling influences on adolescent smoking behavior were investigated. Twenty-eight male and twenty-eight female teenagers were exposed to each of three experimental conditions: a male smoking confederate, a female smoking confederate, and no model present. The order of each of these half-hour long conditions was counterbalanced across subjects. Three aspects of smoking behavior were directly observed: number of cigarettes smoked, number of puffs per cigarette, and duration of each cigarette. The presence of a smoking model increased the number of people who smoked and the number of cigarettes smoked and decreased the number of puffs per cigarette. The sex of the model did not affect the smoking behavior of either male or female subjects. Boys tended to smoke more cigarettes than girls (p less than .06) and boys took significantly more puffs per cigarette than did girls in the model present conditions. The results provide experimental confirmation that modeling influences teenagers' smoking behavior, but the rate of smoking in the absence of a model demonstrates that modeling effects cannot account for the maintenance of smoking.