Main concepts for three different discourse tasks in a large non-clinical sample

Abstract

Background: Semi-spontaneous speech production tasks are commonly elicited to assess discourse ability. When knowledge of a topic, story, or event is shared, it is possible to gauge the informativeness of discourse by evaluating how accurately and completely an individual produces the concepts considered to be essential to the shared topic. This analysis, main concept analysis (MCA), quantifies the degree to which speakers are able to communicate the overall gist of an event. Though MCA is an easy-to-perform, informative, and reliable measure of discourse adequacy, its widespread adoption depends on the development of standardisation and normative reference. Given the large collection of control discourse transcripts available on the AphasiaBank database, it is possible to generate main concept lists based upon a large sample of control speakers and to characterise their performance to establish preliminary normative reference.Aims: The first aim of this study was to develop main concept checklists drawn from a control population for three semi-spontaneous discourse tasks included in the AphasiaBank protocol—a picture sequence narrative (Broken Window), storytelling (Cinderella), and a procedure (Peanut Butter and Jelly). The second aim was to report MCA results for control speakers to provide a normative reference and to stratify the normative information by age.Methods & Procedures: Ninety-two control transcripts, stratified into four age groups (20–39 years; 40–59 years; 60–79 years; 80+ years), were downloaded from the AphasiaBank database. Relevant concepts were identified, and those spoken by at least one-third of the control sample were considered to be main concepts. A multilevel coding system was used to determine the accuracy and completeness of the main concepts produced by control speakers.Outcomes & Results: Main concept checklists for three discourse tasks are provided. Descriptive statistics are reported and examined to assist readers with evaluation of the normative data. No differences between age groups were observed for the Broken Window narrative. For the remaining discourse tasks, the younger half of the sample generally performed differently than the older half of the sample. Additionally, the two younger age groups did not differ significantly from each other, nor did the two older groups.Conclusions: This study provides main concept checklists drawn from a large control sample. Normative information for main concept production is provided for three discourse tasks. The sample distribution is evaluated relative to the normal probability distribution and the sample composition is described, enabling readers to determine the adequacy of normative characteristics of the sample and also the fit between their patient or client and the normative sample. This study also provides information about main concept or event production for four age groups.