Key points are not available for this paper at this time.
Fifteen high school peer tutors and 15 “special friends” interacted daily for a semester with a class of 9 students with severe disabilities. Two measurement systems were employed: an attitude survey and social behavior probes. Social behavior probes were conducted with a familiar student with autism (i.e., a student from the special education class), an unfamiliar student with autism, and an unfamiliar nonhandicapped student. A pretest-posttest experimental design with control group was employed. In general, statistical comparisons of the groups indicated that serving as a peer tutor was equivalent to serving as a special friend across the two measures. However, on one measure (social interaction with an unfamiliar peer with autism) the special friends showed significantly higher levels of social interaction. Both groups interacted more frequently with the familiar student with autism than the unfamiliar student with autism or the unfamiliar nonhandicapped peer. The experimental groups produced substantially longer interactions than the control group.
Haring et al. (Tue,) studied this question.