​​Hosted by SISP student committee, McGill University

& Department of Educational and Counselling Psychology

 
Carolyn Baer is a postdoctoral fellow at the University of California, Berkeley, funded by the Social Sciences and Humanities Research Council of Canada. Her research investigates the cognitive and social tools that children use to learn about the world around them, including feelings of confidence. She holds a PhD in Developmental Psychology from the University of British Columbia.

Learning with Confidence 
by Carolyn Baer, PhD

References

(1)   Bahrami, B., Olsen, K., Latham, P. E., Roepstorff, A., Rees, G., & Frith, C. D. (2010). Optimally interacting minds. Science, 329(5995), 1081-1085.

(2) Baer, C., & Odic, D. (2020). Children flexibly compare their confidence within and across perceptual domains. Developmental Psychology, 56(11), 2095–2101.

(3) Goupil, L., Romand-Monnier, M., & Kouider, S. (2016). Infants ask for help when they know they don’t know. Proceedings of the National Academy of Sciences, 113(13), 3492-3496.

(4) Call, J., & Carpenter, M. (2001). Do apes and children know what they have seen?. Animal Cognition, 3(4), 207-220.

(5) Kirk, C. R., McMillan, N., & Roberts, W. A. (2014). Rats respond for information: Metacognition in a rodent?. Journal of Experimental Psychology: Animal Learning and Cognition, 40(2), 249-259.

(6) Lambert, M. L., & Osvath, M. (2020). Investigating information seeking in ravens (Corvus corax). Animal cognition, 23(4), 671-680.

(7) Rinne, L. F., & Mazzocco, M. M. (2014). Knowing right from wrong in mental arithmetic judgments: Calibration of confidence predicts the development of accuracy. PloS one, 9(7), e98663.

(8) van Loon, M. H., & Roebers, C. M. (2017). Effects of feedback on self‐evaluations and self‐regulation in elementary school. Applied Cognitive Psychology, 31(5), 508-519.

Picture a time you felt unsure. Maybe you were trying out Zoom for the first time or locating an address in a new city. How did it feel? Low confidence can be uncomfortable, a feeling you try to avoid. But sometimes low confidence tells us something really important: we have the opportunity to learn something new.

Feelings of confidence signal the quality of our knowledge. Am I sure that I know her name? Am I likely to get this math problem correct? The confidence we feel in response to these questions gives us a way of measuring how good our memory is for that name, or whether our calculation can be trusted. When our confidence is low, it tells us our knowledge can be improved.

One way to improve our knowledge is to selectively update our low-confidence beliefs. For instance, if a friend is sure about their answer and you’re not sure about yours, it’s very rational to accept your friend’s answer because their confidence is higher. Adults use confidence in this way to make good decisions as groups – sometimes even better decisions than they would have made alone (1). My own work shows that children can compare confidence between two of their own decisions (2), and we think that confidence might help children decide when to trust others, too.

We can also improve our knowledge by employing learning strategies. When you are unsure about an answer, you might ask for help, look it up online, or double-check your work. A key function of confidence is to signal when to use these strategies. A recent paper found that 20-month-old infants asked a caregiver for help finding a hidden object when they felt low confidence, but rarely when they felt high confidence (3). Chimpanzees, rats, and ravens also seek out additional information before making a decision if they feel uncertain about their knowledge (4, 5, 6). Confidence, therefore, seems to be a fundamental signal in the way we learn.

Of course, we aren’t always perfect at using confidence to know when to deploy our strategies. Picture the child who refuses to ask for help when they really need it, or that one student who won’t stop checking their answers at the end of a test. We call this being miscalibrated: when the confidence you convey doesn’t match up with your actual knowledge. Good calibration predicts skill in subjects like math and leads to bigger boosts in learning over time (7).

But good calibration is challenging to learn, and it improves a lot over the preschool and early school years (7, 8). We’re still learning how and why this change happens, but we think one key piece is feedback from others about whether a strategy is needed (8). Feedback can help a learner interpret their confidence: is this feeling of confidence low enough to warrant asking for help, or might I know this content well enough already?

This research suggests that even young learners experience confidence and use it to guide their learning. But this skill isn’t perfect. With some feedback, learners can calibrate their confidence effectively and treat feelings of low confidence as opportunities to learn something new.