Using a Linguistic Analysis Tool to Detect Deception
Detecting deception is important, yet accuracy rates remain low (e.g., DePaulo & Friedman, 1998). Pennebaker et al. (2003) suggest a linguistic analysis tool may be able to detect deception because people use a different linguistic style when telling the truth than when lying. For example, the anxiety experienced by liars may “leak” into their words. The cognitive resources devoted to the lie will be taken from the message. Newman et al. (2003) found support for these hypotheses using the Linguistic Inquiry Word Count (LIWC). Dzindolet and Pierce (2004a) found the LIWC was useful in detecting deception among participants discussing music preferences. This study expands their work to include other topics. Results from the 2 (topic: movie or television) x 2 (topic importance: high or low) x 2 (communication type: lie or truth) design indicated that linguistic analysis tools may be useful in detecting deception across a variety of topics.