1. Spencer, Elizabeth
  2. Bryant, Lucy
  3. Colyvas, Kim


Variability is common in language sample analysis (LSA), arising from personal factors such as age or level of education, or from factors within the text such as its length and purpose. Variability can affect interpretation of results in clinical practice and research studies, as well as the ability to detect change in individuals over time. This article focuses on sample length and time-based variability in the LSA literature and how it has been addressed through a scoping review. We then propose a method for estimating the effect of this common source of variability to allow determination of reliable change in individuals over time. Although some sources of variability are acknowledged in the research literature and clinical evidence-based practice, there has been no consistent method to account for these. The proposed method we present offers a means to address text length and time-based variability and materials and examples to facilitate its implementation in future studies and practice.