| Individuals with autism spectrum disorders display marked difficulties related to social functioning. These difficulties are often sources of persistent distress in addition to being associated with long-term negative outcomes (Greene et al., 1999; Chamak & Bonniau, 2016). Indeed, the severity of social deficits associated with autism have been found to be one of the strongest predictors of long-term outcomes (Howlin, Moss, Savage, & Rutter, 2013). Fortunately, behavioral interventions have been highly effective in teaching a variety of pro-social skills to individuals with autism spectrum disorders (Koegel & Frea, 1993; Pickles et al., 2016). For example, Leaf et al. (2012) used modified behavioral skills training to teach individuals with autism to engage in social skills including: greeting others, offering assistance, giving compliments, and losing graciously. Behavioral interventions often utilize ratio schedules to increase target responding across a variety of behavioral topographies including social interaction (Allen, Hart, Buell, Harris, & Wolf, 1964; Ferster & Skinner, 1957). However, social skills are unique in that they are often 1) highly contextual and excessive responding may be just as socially undesirable as deficient or absent responding and 2) taught in group settings wherein reinforcement contingencies are applied uniformly to students with idiosyncratic behavioral repertoires. Interventions that promote indiscriminately high rates of responding may have unintended effects of increasing skill performance to socially unacceptable levels. For example, a child may emit sportsman-like comments too often or under the wrong conditions (e.g., saying “good job” after every pass in basketball). Furthermore, problem behavior may be reduced using differential reinforcement of alternative behavior (DRA) only to be replaced by another problem (e.g., excessive levels of the alternative response). For this reason, added schedule requirements may be necessary to facilitate socially appropriate skill development. In the context of a group contingency, reinforcement schedules favoring performance of pro-social skills are typically applied to children who exhibit deficits in those areas as well as children whose responding is within normal limits. Such applications have repeatedly been shown to be effective (Stage & Quiroz, 1997; Theodore, Bray, & Kehle, 2004) and to minimize the effort required of interventionists (Cooper, Heron, & Heward, 2007; Davis & Blankenship, 1996). However, given individual differences in baseline responding and/or motivation, simple group contingencies such as DRA may be associated with unacceptably high rates of pro-social responding (e.g., appropriate touching) for some group members. |