/ Child Psychol. Psychiat. Vol. 36. No. 6, pp. 965-976, 1995 Elsevier Science Ltd © 1995 Association for Child Psychology and Psychiatry Printed in Great Britain. All rights reserved 0021-9630/95 $9.50 + 0.00 0021-9630<95)00016-X Sibling Interaction of Children with Learning Disabilities: A Comparison of Autism and Down's Syndrome Fiona Knott,* Charlie Lewisf and Tim Williams* Abstract—Two potentially contrasting hypotheses can be generated about sibling interactions involving a child with Down's syndrome or autism. Research on siblings would predict that learning disabled children adopt responsive roles. Studies of children with autism would predict impovedshed interactions. Home observations were conducted on 30 sibling pairs involving children with autism or Down's syndrome. Both hypotheses were partially supported. All learning disabled children engaged in frequent bouts of interaction, usually directed by their sibling. While children with autism engaged in fewer bouts and imitated less, they did reciprocate their siblings' initiations. Sibling encounters provide a unique opportunity for such children to learn about social relationships. Keywords: Siblings, autism, Down's syndrome, social interaction Introduction Sibling interactions play an important part in the social life of a child, especially before starting school. Dunn (1988) argues that sibling relationships are particularly important in the development of social skills. There is evidence (Abramovitch, Pepler 8c Corter, 1982) that children develop a style of social exchange with their siblings, which they subsequently use with their peers. Interaction between peers is egalitarian, while that between siblings is characterised by role asymmetries (e.g. Brody &: Stoneman, 1986), giving children access to different sets of roles and skills. Cicirelli (1977) demonstrated that children interact differently with their parents than with their siblings. Similarly, Bryant and Crockenburg (1980) found no correlations between mother-child behaviour and older sister-child behaviour on a task, implying that siblings use different interaction styles from their mothers. It is therefore plausible to suggest that siblings have a relationship which might contribute in a unique way to each other's development. Accepted manuscript received 9 September 1995 *University of Reading. fLancaster University. Requests for reprints to: C. N. Lewis, Department of Psychology, Lancaster University LAI 4YE, U.K. 965