An IBM 1050 AV computer system, including a typewriter keyboard, tape recorder, and slides, was used to teach 4-year-olds word and letter recognition. Three studies explored sex differences in relation to differences in socioeconomic status and learning materials. In experiment 1, 16 disadvantaged children explored the keyboard individually for 6 sessions; then a matching task was presented. Eleven keys with words only were exposed. Experiment 2 expanded these procedures, with 20 disadvantaged children participating in 17 sessions. Ten of these subjects were exposed to letters only; 10, to words only. Boys scored significantly higher than girls on the matching task in experiments 1 and 2. No differences were found between word and letter programs. In experiment 3, 15 middle-class boys and girls participated in 10 exploratory sessions followed by a test. No sex differences were found for the middle-class children. However, the letter program was easier for them. The authors concluded that the computer system used did not react fast enough to hold the attention of the children but that a computer-assisted instructional approach which provides for gross motor responses in lieu of verbal ones is well-suited to disadvantaged children, especially boys. References are included. (CM)