Disrupting the bias to Believe: How 3-year-olds Process Deceptive Testimony from a "Broken" Human

• In Event: Poster Session 13
  In Poster Session: PS 13 Topic - Social Cognition

Abstract

Young children have difficulty correctly interpreting deception (Mascaro & Sperber, 2009). Previous research has suggested that source type (Heyman et al., 2013) and access to conflicting first-hand knowledge (Jaswal, 2010) influence children’s ability to interpret deception. In a recent study using a searching task, Palmquist and Kondrad (in prep) found that the bias to believe human testimony was so strong, even children’s own conflicting, first-hand knowledge did not help them search correctly. It was only when “deceptive” testimony came from a “broken” machine – one that always pointed out the incorrect location of a hidden object, just as the deceptive human had – that first-hand knowledge about the location helped children develop a successful search strategy.

In Palmquist and Kondrad (in prep), the deceptive human and the broken machine differed not only in their surface features (one was human and one was a machine), but also in their intentions (the human had the intention to deceive and the machine did not). Therefore, children may have had more difficulty interpreting the deceptive human either because of: 1) an effective superficial heuristic: always trust sources that look like humans, or 2) a cognitive limitation: an inability to integrate deceptive behavior into their default framework of trust. Machines do not have mental lives, or the ability to intend anything at all, which may have made it easier for children to use their own knowledge to generate a successful search strategy.

In the present study, we cross the human with the machine scenario to explore more deeply the course of children’s bias to believe human testimony. Three-year-olds are faced with a human again – but this time, she is “broken”. Children are told that she “had an accident” and therefore, the “part of her that helps her think is broken”. As with the machine in the previous study, because she is “broken”, she will always show children the incorrect location for the hidden object. In other words, this misinformation is coming from a human form as before, but she lacks the intention to deceive much like the machine. Once again, some children will have their own knowledge about the location of the hidden object, and others will have to rely solely on the experimenter’s misinformation.

If children continue to have difficulty interpreting the testimony, it would suggest that their bias to believe has less to do with the intentionality of the source, and more to do with the superficial form the source takes (Table 1). However, if children are better able to interpret the broken human’s testimony, this would suggest their previous difficulty stemmed from an inability to integrate deception into their default framework of trust.

The number of correct searches will be tallied across 8 test trials. The frequency of correct searches will be compared to chance and across groups using chi-square tests and binomial logistic regression. All data will be collected over zoom using pre-recorded video stimuli. The study will be advertised on childrenhelpingscience.com and to families who have expressed interest in participating in research.

Authors

• Carolyn M Palmquist, Amherst College
  Presenting Author

• Robyn Lindsey Kondrad, James Madison University
  Non-Presenting Author