Learning From Surprise: Harnessing a Metacognitive Surprise Signal to Build and Adapt Belief Networks

Abstract

One's level of surprise can be thought of as a metacognitive signal indicating how well one can explain new information. We discuss literature on how this signal can be used adaptively to build, and, when necessary, reorganize belief networks. We present challenges in the use of a surprise signal, such as hindsight bias and the tendency to equate difficulty with implausibility, and point to evidence suggesting that one can overcome these challenges through consideration of alternative outcomes—especially before receiving feedback on actual outcomes—and by calibrating task difficulty with one's knowledge level. As such, we propose that a major function of education—broadly construed as the work of teachers, journalists, parents, etc.—is to assist learners in using their metacognitive surprise signals to facilitate the building and adaptation of belief networks.

Communicating Climate Change Oceanically: Sea Level Rise Information Increases Mitigation, Inundation, and Global Warming Acceptance

“…Because inundation is relevant and/or personally threatening to many Americans, it seems among the likeliest issues useful for influencing behavior/actions (Weber, 2006). Other advantages of sea level rise communication over less salient climate change effects, are clear linkages between oceanic rise and global warming, with scientists proving that warming temperatures have increased hurricanes' strengths and that oceanic expansion increases chronic nuisance flooding frequencies (Milman, 2018)-science demonstrable in simple classroom experiments. The media occasionally, persuasively, state the relationships between climate change and its effects, such as flooding (Olausson, 2009)-and personal experiences with extreme weather events (e.g., flooding) have highlighted climate change for non-victims (Konisky et al, 2016).…”

“…(e.g., Kishore et al, 2018). Beyond storm surges, U.S. flooding frequency from non-storm high tides has doubled in just 30 years, causing human deaths and many billions of dollars in damage, with risks to infrastructure and coastal properties high and soaring (Nicholson-Cole and O'Riordan, 2009;Milman, 2018). Sea level rise's threat clearly impacts America's housing market, with homes more exposed to oceanic rise selling for approximately 7% less than equivalent homes at higher elevations yet equidistant from the beach (Bernstein et al, 2018).…”

Editors’ Introduction and Review: An Appraisal of Surprise: Tracing the Threads That Stitch It Together

“…In the present issue, Loewenstein explores how learning in childhood might build on the kinds of processes that have been found in developmental studies of young children, in that the stories that children learn are optimized to take advantage of children's surprise responses to teach important aspects of culture. In addition, Munnich and Ranney () explore this role of surprise in this issue, laying out a framework of relatively surprising and unsurprising kinds of outcomes that warrant different degrees of memory restructuring. Munnich and Ranney draw on Foster and Keane's () finding that surprise tracks the metacognitive difficulty of explanation—we are more surprised to the extent that we have to explain more about an outcome—to propose that a surprise signal can serve as an index of the amount of memory restructuring (i.e., assimilative vs. accommodative; Piaget, ) required to accurately represent a domain of knowledge.…”

Information Theoretic Characterization of Uncertainty Distinguishes Surprise From Accuracy Signals in the Brain

“…During an initial deliberation phase, for instance, agents form a belief on a decision's outcome, which is graded by confidence (Kepecs and Mainen, 2012). An outcome that challenges beliefs yields surprise (Hsia, 1991;Nour et al, 2018;Munnich and Ranney, 2019). Both confidence and surprise relate to uncertainty in the environment but their characterization remains a topic of debate (Itti and Baldi, 2009;Baldi and Itti, 2010;.…”

“…These two quantities together make up the uncertainty defined in the Free Energy Principle (Friston, 2010), whose minimization is theorized to be the brain's primary purpose (Schwartenbeck et al, 2015) and comprises a compelling theoretical framework for brain function. Questions on the neural characterization of different forms of uncertainty persist for both confidence (Pouget et al, 2016) and surprise (Munnich and Ranney, 2019). Current studies investigating uncertainty in the brain often rely on the notion of a Bayesian brain (Friston, 2012), where a probabilistic model of the world is built (the prior) and subsequently updated (posterior) through repeated interactions with the environment.…”