Experiment 1 investigated the effects of observational training in a social listener reinforcement game on participants' conversational units in non-instructional settings. Experiment 2 tested the effects of multiple exemplar instruction on the development of "empathy." The participants who had reader/writer levels of verbal behavior were selected because they emitted a low number of speaker and listener exchanges with others and did not have empathy as measured in the study. In Experiment 1, data were collected on the number of sequelics, conversational units, "wh" questions, vocal approvals and vocal disapprovals during 5 minute probe sessions after meeting criterion on each objective of the listener reinforcement game. During intervention, the participants were required to engage in a speaker listener exchange through playing a yoked contingency game where they competed with the experimenter to teach several objectives including "I Spy", peer tutoring, textually responding and group instruction. Data showed that the total number of verbal interactions in non-instructional settings significantly increased for both participants. In Experiment 2 multiple exemplar instruction across questions was implemented to teach the participants empathy. The participants responded to 3 questions: "What happened?" "How does the person feel?" and "What could you do to help?" The results showed a significant increase in correct responses to empathy questions. (Contains 8 tables and 6 figures.)