Development and reliability of a quantitative measure of adults' narratives

Abstract

Background: Assessing narrative discourse production in persons with aphasia has long challenged clinicians seeking to improve functional outcomes. Fortunately, the development of single picture or picture sequence stimuli has enabled clinicians to quantify aspects of elicited narrative discourse production in a clinical context. However, also needed for the analysis of the narrative discourse of individuals with aphasia are performance data for adults without brain damage. Such comparative data of both younger and older adults would considerably extend the clinical usefulness of discourse tasks that incorporate picture stimuli. However, elicited narrative discourse samples are only valuable as assessment tasks if the procedure yields samples of similar quality for an individual over time.Aims: The main objectives of this investigation were (a) to characterise the quality of the discourse narratives of non-brain-damaged (NBD) adults, examining the effects of stimulus types on their performance; (b) to compare the proportion of main events conveyed by younger and older NBD adults on the elicited narrative task, and (c) to estimate the test–retest reliability of these tasks with each group of participants.Methods & Procedures: A total of 40 neurologically intact adults were divided into younger (YG; N = 21) and older (OD; N = 19) groups. Participants attended two sessions, 10–20 days apart. Each time, participants viewed two pictures and two picture sequences (Nicholas & Brookshire, 1993) and told what was going on in the pictures. The language samples of each participant were then evaluated for the proportion of main events included, and test–retest reliability was assessed.Outcomes & Results: The YG group conveyed a significantly larger proportion of main events than the OD group. The main effect for picture stimulus was also significant; participants told significantly more main events in response to sequential versus single picture stimuli, regardless of age. Test–retest results yielded strong, positive correlations between sessions for both groups.Conclusions: Our findings suggest that age does influence performance in elicited narrative discourse. The YG group conveyed more causal links and relationships between the events depicted in the pictures than the OD group. Test–retest results indicate that the measure is stable over time for younger and older adults without brain damage.