Global coherence in aphasia during conversation (Leaman & Edmonds, 2020)
datasetposted on 18.06.2020 by Marion C. Leaman, Lisa A. Edmonds
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
Purpose: Global coherence is an essential macrolinguistic discourse skill that speakers use to formulate discourse to convey meaning with maintenance to a topic. When global coherence is poor, the listener’s ability to understand how the discourse makes sense as a whole is diminished. Measures exist to evaluate global coherence in people with aphasia during monologue tasks (e.g., picture description). The aim of the current research is to develop such a measure for unstructured conversation and to explore how global coherence is impacted by aphasia during conversation. A global coherence measure for conversation is required because markedly different cognitive and linguistic demands are made for production of different types of discourse. Thus, a structured monologue measure cannot be used with validity for unstructured conversation. To adequately evaluate global coherence during conversation, a measure specific to the demands of conversation is required.
Method: We adapted the 4-point Global Coherence Scale (Wright & Capilouto, 2012; Wright et al., 2013), a monologue-level measure of global coherence to conversation, resulting in the 4-point Global Coherence Scale in unstructured conversation (GCSconv). We conducted statistical evaluation of the reliability/stability of the 4-point GCSconv in 18 unstructured conversations held by nine people with aphasia. Utterances with low global coherence scores were classified following a recent methodology to describe how breakdown in these utterances contributed to diminished global coherence (Hazamy & Obermeyer, 2019).
Results: The 4-point GCSconv demonstrated excellent inter/intrarater reliability and test–retest stability. Nonspecific language and off-topic comments contributed most frequently to lowered global coherence.
Conclusions: Findings suggest the 4-point GCSconv may be a feasible and reliable measure of global coherence in conversation. This measure adds to a core of emerging reliable discourse measures for conversation. As such, it has potential to inform assessment and treatment of everyday conversation and to investigate the relationship of global coherence in structured monologue and unstructured conversation.
Supplemental Material S1. 4-point GCSconv interrater reliability data for varied sampling methodologies.
Supplemental Material S2. Individual 4-point GCSconv scores.
Supplemental Material S3. Coded conversational excerpt with comments.
Leaman, M. C., & Edmonds, L. A. (2020). Measuring global coherence in people with aphasia during unstructured conversation. American Journal of Speech-Language Pathology. Advance online publication. https://doi.org/10.1044/2020_AJSLP-19-00104
Publisher Note: This article is part of the Special Issue: Select Papers From the 49th Clinical Aphasiology Conference.