Measuring communicative functioning in community-dwelling stroke survivors: Conceptual foundation and item development

Background: Studies employing item response theory methods to evaluate communicative functioning assessment items have found that a broad range of communication tasks and activities may fit a unidimensional measurement model, but that additional item content is needed to extend the range of ability effectively measured by the small subset of items that have been evaluated. Aims: To describe the item identification, evaluation, and development process used to substantiate the content relevance and representativeness of a set of communicative functioning assessment items targeting community-dwelling stroke survivors. Methods & Procedures: Electronic and secondary references were searched to identify assessment tools with item content designed to measure communicative functioning in adults with neurogenic communication disorders. Candidate items were evaluated using face-to-face interviewer-assisted survey groups conducted independently with communicatively impaired stroke survivors (n=59) and their communicative partners (n=61). Web-based surveys were employed to evaluate candidate items from the perspective of practising speech-language pathologists (n=114). Outcomes & Results: A total of 673 items were identified from 33 instruments. A total of 426 met the specified concept definition; 211 were determined to be non-redundant; 166 were identified by key stakeholders as unambiguous, relevant, and moderately to very important to daily functioning. Conclusions: The item pool developed samples a representative range of communication behaviours, activities, and life situations that are relevant to community-dwelling stroke survivors. Further research using item response theory methods is required to substantiate the construct dimensionality and range of ability effectively measured by the item pool, and to evaluate dynamic assessment algorithms designed to minimise response burden.