A Possible Confounding Variable in Different Versions of the “Aggression Machine” When Used in Research on Alcohol

When various versions of the Buss aggression paradigm are used in research concerning alcohol-related aggression, different information about the confederate's degree of intoxication is given to the subject. Sober subjects get the impression that the confederate is sober, whereas intoxicated subjects believe that the confederate is intoxicated. This constitutes a confounding variable making it impossible to attribute conclusively differences among alcohol and control groups on measures of aggression to consumed alcohol. An experiment tested whether this difference in information affects subjects' conception concerning expected level of aggression directed towards them from an opponent and their own level of aggression towards the opponent within the Taylor version of the “aggression machine”. Results clearly indicate that subjects expect intoxicated opponents to be more aggressive. Subjects are also more aggressive toward intoxicated opponents than sober ones. This confirms the importance of holding constant information about the confederate; this has not been done in many relevant studies.