It is assumed that drinking alcohol reduces stress and enhances positive feelings, but studies have not always confirmed this widely-held belief.

One reason could be that the effects of alcohol are often tested on people drinking in isolation rather than in groups. Those studies failed to create a realistic condition for studying the effects alcohol, according to lead author Michael A. Sayette from the University of Pittsburgh.

For this he and his team recruited 720 male and female participants. The participants were divided into groups of three that contained all the gender possibility combinations, from all male to all female groups.

Twenty groups representing each gender composition were randomly assigned to one of three conditions: they drank an alcoholic beverage, a placebo beverage, or a nonalcoholic control beverage.

The alcoholic beverage was vodka and cranberry juice while the placebo beverage was flattened tonic water plus cranberry juice.

To make the participants think they were drinking alcohol, the glasses in the placebo group were smeared with alcohol. In the control condition, the participants knew they were not drinking alcohol.

The participants sat at round tables and consumed their drinks over a 36-minute time span. The group interactions were video taped so researchers could analyze facial expressions and speech behaviors.

When alcohol was consumed, participants spent more time talking to each other, there were more “true” smiles, the group was more likely to smile in unison, group members were less likely to express negative emotions, and they were more likely to indicate that they bonded with other group members.