A Conceptual Model of Real World, High Stakes Deception Detection

There is a strong need for more field studies in deception detection research utilizing objective measures. The study of deception detection is heavily skewed toward laboratory research utilizing university students in mock lie scenarios which are then subjectively scored by humans. This paper introduces a conceptual model that may aid in detecting deception during high stakes interview style communication, utilizing text analytics tools. The model attempts to examine the interactive nature of communication channels within a real world high stakes case study. More specifically it attempts to show that when the source of interactive communication originated from a richer channel in a field setting deception detection accuracy improves even when examined in a less-rich channel. This paper describes a new model and methodological approach utilizing existing text analytics tools that has the potential to positively impact a wide variety of law enforcement agencies and others who conduct high stakes interviews.