Linguistic Inquiry and Word Count (LIWC) software was used to code truthful and deceptive words in prisoners’ natural language. Reality Monitoring (RM) and Newman, Pennebaker, Berry, & Richards’ (NP, 2003) models were used. NP indicates that lies contain fewer self‐references, other references, and exclusive words, and higher numbers of negative emotion and motion words. Higher sensory, spatial, temporal and affective RM terms were predicted for truths, and more cognitive mechanism words were predicted for lies. The RM model’s hit rate was 71.1% and discriminability was 1.11 without spatial words, which were surprisingly higher in lies than in truth statements, and the NP model was 69.7%, d′ = 0.99. The software models were contrasted with humans’ hit rates, and younger prisoners had 71% hits, d′ = 0.22, but older prisoners had 50% hits, d′ = 0.90. The software set unbiased criteria (β), but younger prisoners were biased in setting their criteria when judging statement veracity, β = −0.34. Without other references, found to be higher in truths than in lies, NP classified 59% of statements correctly.
Bond, G. D., & Lee, A. Y. (2005). Language of lies in prison: Linguistic classification of prisoners’ truthful and deceptive natural language. Applied Cognitive Psychology, 19(3), 313-329.