Examines Japanese and American children's styles of processing computerized geometric matrices. Hypothesizes cultural differences for response latencies. Finds no difference in accuracy between the two groups but significant differences in response times for second- and fifth-graders. (MW)