Validation of the Wing Subgroup Questionnaire Using a Concurrent Operants Design

Background: Several subtypes have been proposed to explain some of the heterogeneity in ASD. These include Wing’s subgroup paradigm that classifies individuals as aloof, passive, active but odd, or typically developing based on the frequency and quality of their social interactions (Wing & Gould, 1979). Castelloe and Dawson (1993) adapted Wing’s typology to produce the Wings Subgroup Questionnaire (WSQ), which based subtype on caregiver ratings. Concurrent operants assessments are an example of a preference assessment based on observation of choices. They have frequently been used to assess relative preference for two simultaneously available stimuli (Fisher et al., 1992). Concurrent operants arrangements have been used to assess preference for tangible items (e.g, toys), but have also been extended to assess preference for non-tangible items, such as social attention (Harding, 1999). A concurrent operants assessment may be ideally suited to directly identify subgroups of individuals with ASDs along Wing’s typology.

Objectives: The objective of this study was to directly assess social subtypes using a concurrent operants assessment and compare the results to those of the WSQ for each participant.

Methods: Participants consisted of children with an ASD and their primary caregivers. Each caregiver completed the WSQ. Next, each child participated in a concurrent operants assessment in which preferences for social attention were assessed based on allocation to different sides of a symmetrical room. Throughout the analysis, the participant was free to move about a room that was divided in half by a line on the floor. If the participant stepped onto the same side of the room as the experimenter, they received social attention in the form of descriptive statements and joint play. Whenever the participant was on the opposite side of the room, the experimenter ignored them. Three different conditions were used to evaluate preference for social attention. The experimenter remained on the same side of the room throughout the stay condition. During the chase condition, if the participant chose the non-attention side of the room the experimenter would switch sides and begin to deliver attention after 30 s. Thus, aloof individuals had to continually switch sides to avoid social attention. In the flee condition, if the participant spent 30 s on the same side of the room as the experimenter, the experimenter would switch sides and stop delivering attention. Thus, active but odd or typically developing individuals had to continually switch sides to maintain access to social attention.

Results: All participants choices resulted in the clear identification of a social subtype. However, correspondence between WSQ and concurrent operants assessment was quite low (25%) and statistically insignificant, with approximately half of participants showing greater preference for social attention than was suggested by the WSQ.

Conclusions: Although there was poor correspondence between the concurrent operants assessment and WSQ, this difference could well be explained by a variety of procedural variations, including the use of novel therapists. However, these results raise questions for future research about how best to measure preferences for social attention in those with ASD.