Poster #78 - Developmental Progression in Children’s and Adults’ Interpretation of Confidence Cues

• In Event: Poster Session 08

Abstract

We often think that children absorb everything that they are told, but they are in fact quite selective when learning from others. Indeed, they can rely on many cues, including a person’s level of confidence, to decide whom to learn from (Brosseau-Liard & Poulin-Dubois, 2014; Matsui et al., 2016). Importantly, limited research to date has examined why children use confidence cues. Children could perceive confidence as indicative of a person’s level of knowledge, as a social trait or even as both. To gain a better understanding of children’s interpretation of confidence cues, two experiments investigate the epistemic and social inferences made by adults and children when witnessing a person display either confidence or hesitancy. In Experiment 1, adult participants (N=192, M= 20.28 years, SD=4.39, 151 women) watched a video of a female informant giving geographical directions to a confederate while acting confidently (Confident condition) or hesitantly (Hesitant condition). Participants then answered a series of questions via an online questionnaire to assess their interpretation of confidence (or lack thereof). Participants in the Confident condition, when compared to those in the Hesitant condition, gave higher ratings both when asked to attribute knowledge-related and social traits to the informant. For instance, when asked how smart they thought the informant was on a scale of 1 to 7, participants in the Confident condition gave higher ratings than those in the Hesitant condition (MConfident=5.28, SD=1.37; MHesitant=3.34, SD=1.09, t(188)=10.72, p<.001). A similar effect was found when participants were asked how likable they thought the informant was (MConfident=5.11, SD=1.66; MHesitant=3.73, SD=1.35, t(188)=6.28, p<.001). In Experiment 2 (data collection ongoing), we investigate whether similar trends are noticeable in children ages 7 to 10 using the same stimuli and experimental design as in Experiment 1, while adapting both the number and type of test questions to children’s cognitive level (i.e., fewer questions and simpler vocabulary than in the adult study). We aim to recruit at least 128 participants between the ages of 7 and 10 which is, according to a power analysis, a sample size large enough for independent sample t-tests to have at least 80% power to detect at least a medium effect size. Thus far, the number of participants recruited is too small to present data but the same analyses as in Experiment 1 (e.g., independent sample t-tests) will be conducted once the final sample size is reached. Based on results from the adult experiment, we predict that children will attribute more knowledge-related and positive social traits in the Confident condition than in the Hesitant condition.

Authors

• Aimie-Lee Juteau, University of Ottawa
  Presenting Author

• Carolyn Holmes

• Patricia E Brosseau-Liard, University of Ottawa