A Quantitative Analysis of Language Interventions for Children with Autism

Research and services continue to expand to community-based programs serving individuals diagnosed with autism. A focus of great interest in those efforts is that of language acquisition and functional usage. For the purpose of this evaluation, language acquisition interventions are grouped into two broad categories, contrived and naturalistic. Contrived interventions include those in tightly controlled settings with specific instructional sequences (e.g. discrete trials), whereas naturalistic interventions are defined as less structured and with instructional sequences that change with the child’s interest. This meta-analysis focuses on interventions of spoken language developed for children diagnosed with autism. Contrived teaching approaches were compared to naturalistic approaches during intervention, generalization, and maintenance conditions by calculating the percentage of non-overlapping data (PND). The results illustrate that naturalistic interventions are more effective than contrived interventions for teaching language skills and for maintaining the skills learned. The data also indicate that contrived approaches result in greater generalization than naturalistic approaches. Implications and future directions for the acquisition and generalization of acquired language skills and the importance of treatment integrity are discussed.