Self-reported alcohol expectancies for self and other as a function of behavior type and dosage set.

The hypothesis that expectancies about alcohol's effects on behavior vary as a function of behavior type and dosage set was tested in two self-report studies. In Study 1, 85 male and 88 female college students estimated how a moderate versus high dosage of alcohol would affect three social and three nonsocial behaviors selected from Southwick, Steele, Marlatt, and Lindell's (1981) alcohol expectancy questionnaire. In Study 2, 61 male and 113 female college students estimated how alcohol would affect the same behaviors, but they estimated the effects on another person (male or female) rather than on themselves. Results from both studies indicated that subjects expected a moderate dose of alcohol to have greater effects on social than nonsocial behavior, and they expected a high dose to have greater effects on nonsocial than social behavior. Gender also influenced the expected strength of alcohol effects.