Memory for Conversation in Traumatic Brain Injury: A Feasibility Study and Preliminary Findings

Abstract

Purpose: Despite common clinical complaints about memory for conversation after traumatic brain injury (TBI), the nature and severity of this deficit are unknown. In this research note, we report feasibility and preliminary data from a new conversation memory study protocol.

Method: Participants in this feasibility study were 10 pairs, each including an adult with chronic, moderate-to-severe TBI and their chosen familiar conversation partner. The experiment began with a naturalistic conversation between participants with TBI and their conversation partners. After a filled delay, participants next completed verbal recall for the conversation, which we transcribed and coded for their accuracy relative to the original conversation. Participants also read chosen statements from their original conversation and predicted what each partner would remember in a week. One week later, participants completed a posttest about who said each of the chosen statements, allowing direct comparison to their predictions.

Results: We successfully collected conversation memory data from all 10 pairs, suggesting that this protocol is feasible for future study. In this preliminary sample, people with TBI and their conversation partners did not differ in the accuracy of their recall for the conversation about 20 min after it occurred. When asked to predict their partner's delayed memory, conversation partners were less accurate than participants with TBI because they underestimated how much their partners with TBI would remember.

Conclusion: Measuring memory for conversation in TBI is feasible and may advance the characterization of cognitive-communication impairment in TBI, and its heterogeneity, in everyday contexts.

Supplemental material: https://doi.org/10.23641/asha.25927513.

Figure 1.

Percentage of Judgment of Learning (JOL) statements remembered by each group. Black boxes represent actual performance at the 1-week posttest, and gold boxes represent performance as predicted by the other member of the dyad. For example, the gold box for the CP group represents the percentage of source attributions that partners with TBI predicted that CP participants would remember in 1 week, and the black box represents the percentage of source attributions that CPs did remember at the 1-week posttest. Central lines represent group medians, and points represent individual data points. CP = familiar conversation partners; TBI = participants with traumatic brain injury.