Measuring performance stability in persons with aphasia: identical versus comparable testing forms

Background: Persons with aphasia often show variability in language behavior. It is thus a common practice to evaluate their skills on multiple occasions, for research and clinical purposes. This repeated probing usually relies on identical testing materials, sometimes leading to practice effects. As an alternative, different but comparable testing materials used on successive occasions can be applied. Aims: To evaluate simultaneously performance stability on comparable testing forms, the less common method, in relation to performance stability on identical forms. Methods & Procedures: We tested five persons with non-fluent aphasia using three comparable testing forms that were administered twice. Naming accuracy in an action picture naming task and overall productivity, informativeness, grammaticality, and lexical diversity in two tasks eliciting connected speech (a personal narrative and a picture sequence) were assessed. Outcomes & Results: Pearson’s correlation coefficients were calculated among the three comparable testing forms administered at each testing time. Pearson’s correlation coefficients were also calculated between identical forms administered twice. We then contrasted the correlation ranges across identical and comparable forms. With a few exceptions, performance was highly correlated (r > .80) among identical forms and among comparable forms. Conclusions: Our preliminary findings suggest that the reliability of both testing methods is similar, which may facilitate future use of comparable forms, a currently less common method in aphasia research. This testing approach can be particularly useful where practice effects are likely to occur.