Three studies, using handicapped children, investigated: effects of three different reinforcement contingencies (positive reinforcement, removal of positive reinforcement, and combination of positive reinforcement and removal of positive reinforcement) on a steady-state discrimination task; effect of a conditioned emotional response procedure on a steady-state discrimination task; and effect of a conditioned emotional response procedure on rate of words emitted. Results of the three studies indicated: percent of correct discriminations was highest for removal of positive reinforcement; no significant differences were found in response latencies for any of the three reinforcement conditions; conditioned emotional response procedures had no effect on response latency or percent correct discriminations; conditioned emotional response procedures had no effect on rate of words emitted. Major implications of studies were discussed. (Author)