Cognitive task demands and discourse performance after traumatic brain injury.

The sample was mostly white, from the upper Midwest United States, and mostly had post-secondary education, so the results may not apply to other sociodemographic groups. Raters were undergraduate students and mostly female, so their judgments may not reflect broader social audiences. The TBI group was high functioning, with low-average neuropsychological scores and mostly legal independence, so findings may not apply to people with more severe cognitive impairment. The word-ban condition used word avoidance rather than inhibition of irrelevant thoughts, so it may not match everyday conversation demands. All participants spoke with a scripted female investigator who gave only minimal responses, so partner effects may differ in real conversation. Some participants did not complete all topics because of fatigue, agitation, anxiety, instruction failure, or equipment problems, and excluded data may affect generalization. The study was exploratory and used correlations, so it cannot show that mental state terms or dysfluency cause poor social acceptance.

Do not read the listener ratings as evidence that mental state terms cause social acceptance. The ratings were correlations with transcripts, and the study did not manipulate mental state terms.