Non-transcription analysis of connected speech in mild cognitive impairment using an information unit scoring system

Abstract

Purpose: The purpose of the current study was (1) to compare the performance of individuals with mild cognitive impairment (MCI) and cognitively healthy adults (CHA) on story retelling tasks consisting of familiar and novel stories, and (2) to examine potential utility of information unit scoring by demonstrating concurrent validity of the measure. Method: Fifteen individuals with MCI and 15 age- and education-matched CHA who are monolingual, Korean speakers participated in this study. The task consisted of two different stories: (1) familiar story and (2) novel story. Each story was presented auditorily with sequential pictures. All participants were required to retell as much of the story as they could remember after listening to recorded audio files. Language samples were quantified using Information Unit (IU) checklists. Concurrent validity was computed by correlating IU measures (percent of information units [%IU] and percent of information units per minute [%IU/Min] and other linguistic measures (e.g., correct information units [CIU]). Result: There were greater differences between the two groups in the familiar story than the novel story. Correlations for IU measures in the familiar story were significant and moderate to high, ranging from 0.644 to 893 for the MCI group and from .582 to .745 for the CHA group. In the novel story, one moderate correlation was found between the %IU/Min to %CIU/Min. Conclusion: This evidence supports that the IU scoring system serves as a valid and efficient tool to analyze connected speech in clinical settings. © 2021