Gender and social comparison effects in computer-based problem solving

Gender differences in relation to school children's learning with computers are frequently attributed to a tendency for boys to dominate computer resources in mixed sex settings. However, the evidence relating to children's performance with computers in mixed sex groups is conflicting. This paper reports two experimental studies in which 11- to 12-year-olds worked on a computer-based problem solving task. In the first, 62 children worked in either same or mixed sex dyads, but each child had her or his own computer, and no verbal interaction was allowed. Boys out-performed girls overall, with sex differences becoming significantly more polarised in the mixed sex dyads. The second study involved 96 children, with individual pre- and post-tests, and compared co-action dyads (as in the first study) with interaction pairs, in which the pair members worked together at a single computer, with no restriction on interaction. The polarisation of sex differences in the mixed sex dyads was once again found in the co-action condition, but not in the interaction condition. Results are interpreted in terms of processes of social comparison, which appear to be more potent in this situation than any straightforward domination of resources.