Determining the efficacy of a usage-based language intervention in the early years : a non-randomised pilot stud

Abstract

Introduction Language skills are crucial for positive social, emotional, academic and economic outcomes across the lifespan. Whilst most children acquire language relatively easily, a number of children experience difficulties. Early intervention can mitigate the risks associated with poor language skills, and the importance of evidence-based intervention is widely recognised by Speech and Language Therapists. Despite this, there are few robustly evaluated language interventions for young children. Building Early Sentences Therapy (BEST) is a complex language intervention designed to support children to understand and produce two, three and four clause-element sentences (McKean et al., 2013). This thesis presents a non-randomised quasi-experimental study of BEST with sign, BEST without sign and Treatment as Usual with four key goals: 1) to evaluate the efficacy of BEST, 2) to assess the use of sign as an ‘active ingredient’, 3) determine the effect sizes from goals 1 and 2, and 4) to inform further evaluation. Methods Schools were assigned to three treatment arms matched with respect to classroom oral language environment and socioeconomic status. Participants were 3;5-4;5-year-old children with expressive, receptive or mixed language difficulties below the 16th centile. The intervention was 16 sessions of BEST with or BEST without sign. The TAU group received their usual classroom provision. Children’s outcomes on production, comprehension, content and morphology on targeted and non targeted language structures were measured. Results Visual trends showed improvement for children receiving BEST. There was evidence for the efficacy of BEST compared to TAU for Production, Content and Morphology outcomes. There was also evidence that BEST with Sign improved outcomes further for Production and Sentence Morphology outcomes. Content Analysis of teacher feedback led to the generation of recommendations for future studies evaluating BEST. Discussion Children receiving BEST make progress with production, content and morphology outcomes. The role of sign in further supporting progress is positively indicated. A further trial of BEST is warranted.