Deception detection based on the linguistic style of honest and dishonest stories

Abstract

Purpose This study aims to describe and analyze the differences in the linguistic styles of honest and dishonest stories. Design/methodology/approach This paper uses a descriptive study with a multivariate analysis of linguistic categories according to the story. The research analyzed 37 honest stories and 15 dishonest stories produced during actual legal proceedings through software Linguistic Inquiry and Word Count (LIWC). Findings The authors find that individuals who engage in deception use a different number of words when they narrate facts. The results suggest a need for additional investigation of the linguistic style approach because of its high applicability and detection accuracy. This approach should be complemented by other types of verbal, nonverbal and psychophysiological deception detection techniques. Research limitations/implications Among the limitations, the authors consider length of the stories should be considered and scarce scientific literature in Spanish to compare with outcomes in English. Practical implications This research highlights the relevance to include linguistic style in real contexts to differentiate honest and dishonest stories due to objectivity and agility to implement. Social implications Understanding deception as a social behaviour and its psychological processes associated are elements that contribute to people and justice to comprehend it. Originality/value Analyzing real statements and discriminate differences in linguistic style, contribute to understand deeply this important behaviour to propose new methodologies and theories to explain it.