Are You Lying to Me? Exploring Children's Nonverbal Cues to Deception

Are You Lying to Me? Exploring Children’s Nonverbal Cues to Deception Mariana Serras Pereira (M.SerrasPereira@uvt.nl) Eric Postma (E.O.Postma@uvt.nl) Suleman Shahid (S.Shahid@uvt.nl) Marc Swerts (M.G.J.Swerts@uvt.nl) Tilburg center for Cognition and Communication (TiCC), Tilburg University, The Netherlands important that the lies that are investigated are natural and spontaneous, and in that way representative of the behaviour children exhibit in their normal social contexts, which would render acted versions of lies less suitable for research purposes. To introduce our own approach to detecting nonverbal cues in children’s expressions, we first describe previous studies into deceptive behavior of children, then review previous findings of nonverbal correlates of lying behavior and then say a few words about methods to (automatically) detect lies. We then embark on a description of our own study, which consists of a specific elicitation paradigm, a perception study, and a variety of (automated) detection methods. Abstract The present study investigates how easily it can be detected whether a child is being truthful or not, and explores the cue validity of a child’s body movement for such type of classification. To achieve this, we introduce a combination of methods, in particular a perception test, and an automated body movement analysis. Film fragments from truthful and deceptive children were shown to human judges who were given the task to decide whether the recorded child was being truthful or not. Results reveal that judges are able to reliably and accurately distinguish truthful clips from lying clips. The automated movement analysis revealed a positive correlation between the amount of movement in a child and the perception of lies, i.e., the more movement the children exhibited during a clip, the higher the chance that the clip was perceived as a lie. Children’s Lying Behaviour Keywords: Lying; Deception; Children; Nonverbal Cues; Body Movement; Automatic Analysis. Introduction A question which has intrigued many generations of researchers is whether and how one is able to detect if a conversation partner is being truthful about the things he or she is claiming, or not. Apart from criminal and juridical reasons, this question has been deemed relevant for educational and developmental purposes as well. Accordingly, this has led to a series of studies into child- specific aspects of deceptive behavior (Fu et al., 2012; Ruffman et al., 2012; Talwar C Talwar & Lee, 2002a, 2002b). Obviously, one could think of many situations in which parents, caregivers, or teachers would find it useful to know whether or not a specific child is trying to deceive them, even when these may mostly relate to innocent issues like a broken window, a stolen cookie or a fight with another child. In particular, there has been a specific interest in nonverbal features (such as specific facial expressions or eye gaze patterns) that children could possibly display when they are telling a lie. However, as we will show below, in a review of the literature, the evidence regarding the usefulness of such nonverbal features as markers of deceptive behavior is quite inconclusive. The variability in reported results could partly be due to (1) the kinds of features that have been investigated in terms of their cue value and (2) the techniques that have been used to detect such features. Moreover, it would also seem Previous research (Lewis, Stanger, & Sullivan, 1989; Talwar & Lee, 2002b) suggests that children are quite good manipulators of their nonverbal behavior when lying, which makes the discrimination between truth-tellers and lie-tellers very difficult to accomplish. Most studies report that children’s lies can be detected around chance level, comparable to what has been claimed for adults (Edelstein, Luten, Ekman, & Goodman, 2006; Bond & DePaulo, There is evidence that children start lying from a very young age as early as 2½ years old (Newton, Reddy, & Bull, 2000), and according to Talwar & Lee (2002b) lie- tellers between 3 and 7 years old are almost indistinguishable from truth-tellers. Nevertheless, some studies suggest that lie-tellers tend to exhibit slightly more positive nonverbal behaviours, such as smiles, relaxed and confident facial expressions, and a positive tone of voice, which is also congruent with the findings from Lewis, Stranger & Sullivan (1989). In addition, earlier work also suggests (e.g. Feldman, 1979) that children have a poor control on their nonverbal behavior. The extent to which children display nonverbal cues could be related to the kind of lie and to the circumstances under which these are told. In earlier work ( Swerts, 2012; Swerts, van Doorenmalen, & Verhoofstad, 2013), it was reported that children tend to leak more cues to deception when they are more aware of their deceptive attempt: for example, children’s second attempts to lie (after having been told to repeat a previous lie) revealed more nonverbal cues in their