About this srcd poster session

---

Panel information
Panel 4. Cognitive Processes

Abstract

---

People, especially children, often indicate higher confidence than their accuracy would warrant (Destan & Roebers, 2015; Mazancieux et al., 2020). A focus of research has therefore been to recalibrate confidence and help people properly match their confidence to their accuracy. Many efforts to reduce overconfidence center around providing feedback about the accuracy of judgments or the accuracy of confidence ratings of those judgments, with varied success (Buehler et al., 2023; Carpenter et al., 2019; Rouy et al., 2022; van Loon & Roebers, 2020). Here, we investigate another potential route: providing demonstrations of success or failure.
A demonstration of success/failure is supported by two theories of how confidence representations are formed. First is a Bayesian account, in which participants have a prior expectation about the chances of success at the task (Meyniel et al., 2015). Consider the task of throwing a dart into a bullseye. A demonstration might cause participants to adjust their prior expectations up (if viewing a success) or down (if viewing a failure). Second is through fluency, a feeling of ease associated with processing the given task (Alter & Oppenheimer, 2009). A demonstration should increase the fluency people experience when they imagine throwing darts. In fact, two aspects of the task might be made more fluent. One is fluency about the overall activity (throwing darts), in which case participants should always be more confident after viewing the video, regardless of success or failure. But a successful vs failed outcome could moderate this more general fluency, in which case participants should be more confident after watching a successful demonstration and less confident after watching a failed demonstration.
To test these predictions, we adapted a method that increases confidence in adults (Jordan et al., 2022; Kardas & O’Brien, 2018). Two-hundred and thirty-seven participants (118 children aged 5-12 years and 119 undergraduate adults, all recruited from a suburban Pacific Northwestern area) viewed 3 videos of “trick shots” including darts, ring toss, and bowling. In a within-subjects design (counterbalanced between-subjects), each participant viewed one clip of 5 Successful attempts, one clip of 5 Failed attempts, and one clip of 5 Control attempts (where no outcome was revealed). After each clip, participants rated how many of 10 attempts at that trick shot they could ‘win’.
Our preregistered Age*Outcome mixed-model regression revealed a significant main effect for Outcome and no other effects, b = 0.74, t(472) = 6.93, p < .001. Children and adults anticipated more successes after viewing Successful attempts relative to Failed (paired t(236) = 6.58, p < .001) and Incomplete attempts (paired t(236) = 4.50, p < .001). Children and adults also gave lower ratings following Failed attempts relative to Incomplete ones (paired t(236) = 2.56, p = .011).
These findings show that children and adults use demonstrations of success and failure to increase and decrease their own confidence, respectively. This method is promising for those interested in manipulating confidence and is consistent with two major theories of how confidence representations form.

Author information

Author	Role
Carolyn Baer, Ph.D., Algoma University	Presenting author
Maryanne Garry, University of Waikato	Non-presenting author
Daniel M. Bernstein, Kwantlen Polytechnic University	Non-presenting author

⇦ Back to session