Forensic statement analysis (FSA) research has shown that statements provided by truthful persons may differ from those proffered by individuals who are lying. However, these data are from low stress laboratory study environments that do not compare to situations of stress experienced by people experiencing legal jeopardy or threatening events. As such, our current understanding about how well this may apply to real world situations is limited. This study tested whether speech content of genuine eyewitness accounts about exposure to a highly stressful, personally relevant event differed from deceptive accounts.

Military personnel were randomized to Genuine or Deceptive Eyewitnesses groups. Genuine eyewitness reported truthfully about their exposure to interrogation stress associated with military survival school training; Deceptive eyewitnesses studied transcripts of genuine eyewitnesses for 24 hours prior to being interviewed. Cognitive Interviews of all were recorded, transcribed and assessed by FSA raters blind to the status of participants. The accounts of genuine eyewitnesses contained more external and contextual referents, more unique words, a greater total word count and lower type-token ratio (TTR) than did deceptive accounts.

The classification accuracy when using FSA techniques on Cognitive Interview elicited eyewitness accounts was 82%; FSA methods may be effective in determining whether eye witness accounts about real world, high-stress events are genuine; these data have relevance to professionals working with asylum seekers, or professionals in law enforcement, security and in criminal justice.