The generalizability of social loafing (the phenomenon that people exert less effort when performing in groups than when performing individually) was tested, using Jordanian college students. Subjects in groups of four were asked to think of as many ideas as possible to explain their opposition to a proposed comprehensive examination. The degree of task involvement and output identifiability were manipulated. Consistent with previous findings in the United States, the Jordanians' performance was not as good when they worked collectively as it was when they worked alone. The social loafing effect was eliminated when the Jordanians performed highly involved tasks, regardless of the identifiability of their output.