Thursday, May 20, 2010
Franklin Hall B Level 4 (Philadelphia Marriott Downtown)
3:00 PM
Background: Diagnosis of an Autism Spectrum Disorder (ASD) relies on use of the ADOS (Lord et al., 2003) and ADI-R (Rutter et al., 2003) to assess behaviors associated with DSMIVTR criteria. Criteria also requires an assessment of a child’s “development of peer relationships” (APA, 2000), yet there are no behavioral observations dedicated to this observation. Most evaluators rely on parental reports of children’s social interest, through the ADI-R and/or Vineland (Sparrow et al., 2005). Although parent report can be reliable for out-of-school settings, gathering behavioral information during peer social interactions is lacking. Wing and Gould (1979) originally classified children with autism by their interest in other people, yielding three classifications (passive, avoidant, active but odd). A modified assessment using Wing and Gould’s classifications targeting peer interest is needed.
Objectives: To describe a standard protocol for assessing peer-related interest in children with autism as Avoidance/Escape, Toy-Related Interest, Active but Odd.
Methods: Participants included nine preschool-aged children with autism and nine typical preschoolers. Videotaped probes were obtained during various arrangements of play activities. Condition 1: ALONE offered highly preferred toys, no requirements for peer proximity, and no prompts/rewards for social interaction. Condition 2: OUTDOOR RECESS assessed naturally occurring peer proximity and social contacts during outdoor recess. During Condition 3: PRIMED PEER PROXIMITY children were redirected to stay within 3-feet of typical peers, but received no prompts/rewards for social interactions. During Condition 4: PRIMED SOCIAL CONTACT children were required to stay within 3-feet of typical peers, and typical peers were rewarded for frequent social bids. Peer interest was assessed via (a) proximity to peers across conditions, (b) and conditional probabilities.
Results: Differences between children with and without autism were evident for the percent of intervals children watched or interacted with peers across Conditions 2, 3, and 4. Classifications of social interest were accurate for 8 of 9 children with autism as validated by clinical judgment and more extensive data samples. Three children with autism were classified as AVOIDANT due to displayed levels of peer proximity below the mean and bottom range for both typical children and other children with autism. Three children with autism were classified as having a TOY RELATED INTEREST based on probabilities for “Given child focus on peer, what is the probability child focus on a toy?” approximated typical children and far exceed the range of children with autism. Three children with autism were classified as having an ACTIVE BUT ODD based “Given a social bid from a peer, what is the probability of a verbal response given?”, and on peer responsiveness far exceeding mean for children with autism and lower than normal ongoing rates of social bids from peers.
Conclusions: Peer-related social interest was accurately classified in 89% of children based on data gathered during probe conditions. These classifications suggest that differing peer-related treatment protocols may be effective in remediating these social difficulties. Combined with the ADOS and ADI-R, this standard protocol for effective assessment of peer-related social behavior may be used to accurately identify children with an ASD.
Objectives: To describe a standard protocol for assessing peer-related interest in children with autism as Avoidance/Escape, Toy-Related Interest, Active but Odd.
Methods: Participants included nine preschool-aged children with autism and nine typical preschoolers. Videotaped probes were obtained during various arrangements of play activities. Condition 1: ALONE offered highly preferred toys, no requirements for peer proximity, and no prompts/rewards for social interaction. Condition 2: OUTDOOR RECESS assessed naturally occurring peer proximity and social contacts during outdoor recess. During Condition 3: PRIMED PEER PROXIMITY children were redirected to stay within 3-feet of typical peers, but received no prompts/rewards for social interactions. During Condition 4: PRIMED SOCIAL CONTACT children were required to stay within 3-feet of typical peers, and typical peers were rewarded for frequent social bids. Peer interest was assessed via (a) proximity to peers across conditions, (b) and conditional probabilities.
Results: Differences between children with and without autism were evident for the percent of intervals children watched or interacted with peers across Conditions 2, 3, and 4. Classifications of social interest were accurate for 8 of 9 children with autism as validated by clinical judgment and more extensive data samples. Three children with autism were classified as AVOIDANT due to displayed levels of peer proximity below the mean and bottom range for both typical children and other children with autism. Three children with autism were classified as having a TOY RELATED INTEREST based on probabilities for “Given child focus on peer, what is the probability child focus on a toy?” approximated typical children and far exceed the range of children with autism. Three children with autism were classified as having an ACTIVE BUT ODD based “Given a social bid from a peer, what is the probability of a verbal response given?”, and on peer responsiveness far exceeding mean for children with autism and lower than normal ongoing rates of social bids from peers.
Conclusions: Peer-related social interest was accurately classified in 89% of children based on data gathered during probe conditions. These classifications suggest that differing peer-related treatment protocols may be effective in remediating these social difficulties. Combined with the ADOS and ADI-R, this standard protocol for effective assessment of peer-related social behavior may be used to accurately identify children with an ASD.