Limited Relations Between Varied Measures of Deception and Theory of Mind in Early Childhood

• In Event: Poster Session 06
  In Poster Session: PS 06 Topic - Social Cognition

Abstract

Two important and early-emerging components of social cognition are theory of mind (ToM) and deception. Previous research has found correlational and causal links between first order false belief and antisocial lying, or lying to benefit oneself (e.g., Ding et al., 2015; Evans et al., 2013; Talwar & Lee, 2008), but ToM and deception span many additional domains (e.g., white lies, visual ToM tasks). Given mounting evidence for the fractionated nature of ToM (Warnell & Redcay, 2019) and initial work showing inconsistent relations between ToM and different types of lies (Hsu & Cheung 2013; Sai et al., 2020), systematically examining relations between multiple ToM and deception measures will offer important insight into the underlying structure of social cognition.

Fifty-nine children (32 males) aged 4-7 (mean=6.00y) completed a theory of mind battery (Table 1) that assessed (1) false belief understanding, (2) understanding differences between real and displayed emotion, (3) ability to decipher mental states from photographs of the eyes, and (4) understanding of how others might interpret an ambiguous picture. Additionally, children completed a prosocial, or white, lying task in which they gave feedback to an artist about a bad drawing (Warneken & Orlins, 2015). Finally, children completed three instrumental lying games (based on Ding et al., 2015): (1) a self-benefit game in which the child’s lie would result in them earning a sticker, (2) an other-benefit game where the child’s lie would result in another earning a sticker, and (3) a spiteful lying game, where a lie would result in no one getting the sticker. Children completed multiple trials of each ToM and deception measure.

Each ToM and deception task produced a wide range of performance, indicating that children did not show an all-or-none mastery of these social cognitive skills. Next, we examined relations between tasks. Controlling for age and verbal IQ, we found minimal relations between our measures (Table 2). Within our ToM battery, we found no significant relations between tasks, with correlation coefficients similar in magnitude to previous research. For the deception measures, rates of instrumental lying were significantly related across contexts, but none of the instrumental lying tasks were correlated with telling prosocial lies. Additionally, none of our ToM measures significantly predicted performance on any deceptive tasks.

Our results replicate recent findings of minimal correlations between ToM measures in early childhood and extend these findings to the domain of deception. The wide range of performance on our measures underscores that children often display partial and fragmented mastery of social cognitive skills. Further, we find only weak links between ToM and lie telling. This null finding compared to past work may be due to differences in our deception measures or to our slightly older sample; different factors may underlie the early emergence of simple deception versus its ultimate implementation across complex contexts. Overall, our findings underscore the importance of continued research that systematically targets how broad social cognitive constructs operate under specific instantiations. Such research will advance our understanding of individual differences in real-world child behavior.

Authors

• Callie Brooke De La Cerda, University of Texas at Dallas
  Presenting Author

• Hailey Thomas, Texas State University
  Non-Presenting Author

• Katherine Rice Warnell, Texas State University
  Non-Presenting Author