Differences in the communicative intent among speakers in false and (presumed) true confessions

At my upcoming poster presentation, I’ll share early results from a project examining communicative intent in police interviews with false and (presumed) true confessions. Instead of focusing only on what people say, I analyse how they speak—looking at patterns of questions, answers, listener signals, and other discourse moves across complete interviews.
Using a combination of codebook and large language models, I map thousands of interview units into different communicative functions. This allows us to compare the “profiles” of false versus (presumed) true confessions without relying on constructs such as appropriateness.
If you’re interested in investigative interviewing, wrongful convictions, or applied NLP, I’d be happy to see you there.
See references of the poster here

Title: Differences in the communicative intent among speakers in false and (presumed) true confessions
When/where: Downtown Campus, Ontario Tech University (Charles, 61, Oshawa, Canada) Time: 5 p.m.