Abstract
The impact of grouping, learning handicap, locus of control, and self esteem on students' performance on a math problem-solving program was assessed in two studies. Outcome measures included: 1) the level of difficulty at which students selected to work, 2) use of a program help command, and 3) response accuracy with and without help. Despite indication from the non-computer instructional literature that grouping would facilitate problem-solving skills for some students, neither problem selections nor performance accuracy varied as a function of group configuration. Personal characteristics influenced problem selections, help seeking, and accuracy in anticipated directions but not consistently. More consistent patterns were noted as a function of initial student competency in math.