The Autism Society Event and Education Recordings Archive

In the context of the general call for evidence-based practices and the increasing number of interventions for students with autism spectrum disorders (ASD), it becomes important to critically evaluate those interventions (Heflin & Simpson, 1998). At times, however, the dialogue between research and practice becomes complicated. A number of interventions undergo a rigorous empirical evaluation and are established as efficacious under laboratory conditions, but may be ineffective when used in naturalistic contexts. On the other hand, many interventions emerge from clinical practice and require scientific validation. Social Stories™ for students with ASD provide an excellent illustration of the second scenario. While this intervention has a strong practical rationale and may be appealing to many parents and practitioners, some researchers have argued that its scientific base is yet to be established (Sansosti, Powell-Smith, & Kincaid, 2004).

Social Stories are individualized short stories written by parents or teachers of students with ASD, with the goal of objectively sharing important social information with individuals with ASD (Gray, 2004). The intervention addresses one of the core impairments in ASD, difficulties in the social domain, by describing situations or concepts that may be confusing or difficult for students. A previous meta-analysis of research (Reynhout & Carter, 2006) documented the questionable effectiveness of Social Stories, with highly variable individual outcomes. However, the specific sources of the high variability of intervention outcomes remain unexplored. Furthermore, the use of Social Stories and the reasons for their popularity in school settings have never been documented. Therefore, a systematic examination of the use and effectiveness of Social Stories in research and educational practice is needed.

The goal of this presentation is to describe and discuss the results of two related studies, a meta-analysis of single-subject research and a web-based survey conducted with teachers of students with ASD. The research questions addressed by both of the studies were:

1. What is the overall effectiveness of Social Stories?

2. How are Social Stories constructed and used?

3. What are the critical participant and intervention variables that moderate the effectiveness of Social Stories?

Percentage of non-overlapping data (PND) (Scruggs & Mastropieri, 1998) was calculated for the total intervention, each study and each participant in the meta-analysis. A stringent set of inclusion criteria was used; studies using non-experimental AB designs, treatment packages, those involving participants without a diagnosis of ASD and those with floor/ceiling effects in baselines were excluded. As a result, a total of 18 studies (15 published articles and 3 dissertations) published between 2002 and 2009 were included in the meta-analysis. The survey data were collected in a 36-item web-based questionnaire and analyzed using descriptive statistics. The final sample consisted of a total of 42 teachers working with students with ASD.

The use and effectiveness of Social Story interventions was compared and contrasted across the two studies. Results of the meta-analysis suggested that Social Stories had low to questionable overall effectiveness, as indicated by the PND score of 62% (range: 11-100%) obtained for the intervention. However, the majority of teachers perceived them as highly effective, practical and efficient. The meta-analysis demonstrated larger effects of Social Stories on the reduction of challenging behaviors than on the increase in social skills. In contrast, teachers perceived them as most effective when used to address transition difficulties of students with ASD. Both studies suggested that the use of Social Stories in special education settings with teachers as agents of intervention was the most common. However, results of the meta-analysis indicated that the use of Social Stories in general education classrooms with students as their own intervention agents was associated with higher treatment effectiveness. The use of Functional Behavioral Assessment (FBA) results to inform Social Story interventions was associated with improved outcomes, both in the meta-analysis and in the teacher survey. Finally, younger and higher functioning participants (e.g., students with better communication and social skills) seemed to benefit from the intervention to a greater extent than older students with lower levels of skill development.

The role of additional variables of interest (e.g., format of Social Stories, use of comprehension checks, length and intensity of the intervention, participant reading skills) will be discussed. A discussion of future directions and implications for research and practice will also be provided.