November 14, 2019 · cognitive science of learning theory

Surprise

Did you know that left-handed people die on average 9 years younger than right-handed people? Surprising, right? More on this later, I promise.

How might surprise affect learning? Intuition tells us that when we are surprised we are more engaged ("Whoa! I'm going to pay attention to that thing that is surprising!"), we are motivated ("What could explain that surprising thing?"), and we are possibly delighted ("Yay! New and interesting things are fun!"). Surprise should jolt us into a meta-cognitive state ("How could that be true?!"). Each of these psychological/emotional/cognitive states has clear intuitive links to learning.

It turns out, our intuition is correct. Let's define our theoretical terms: Reisenzen, et al (2017) note that surprise "is evoked by unexpected (schema-discrepant) events". I will pause for a moment to savor a delicious sentence in the abstract – such lovely, precise scientific language:

Unexpected events cause an automatic interruption of ongoing mental processes that is followed by an attentional shift and attentional binding to the events, which is often followed by causal and other event analysis processes and by schema revision.

In other words, surprising events cause us to take notice and to search for an explanation that fits the surprising thing into the way we see the world.

Foster and Keane (2018) maintain that surprise is a "process of 'sense-finding' or 'explanation-finding'" that "increases with the explanatory work required to resolve it." That study found that

surprising outcomes (less‐known outcomes) that are more difficult to explain are recalled more accurately than less‐surprising outcomes that require little (known outcomes) or no explanation (normal).

Put simply, if we encounter something surprising, we are motivated to explain it, and the more difficult it is to explain, the more surprising it is and the more memorable it is as a result. All this sounds like educational gold. Present students with surprising things so that they will be intrinsically motivated to discover an explanation for it, thus driving engagement and "schema revision."

Examples of surprising things in different learning domains:

The positive implications are obvious. We can use surprise as a tool for boosting learning. Are there reasons to only do this sparingly, or to do it cautiously?

Let's revisit the surprising fact I offered at the beginning of this piece: Left-handed people die, on average, 9 years sooner than right-handed people. This should strike almost anyone as surprising, since we typically don't associate handedness with such extreme life expectancy differences. It also works against what we commonly understand to be the causes of mortality. It generates the appropriate metacognitive response: Is this really true? What would explain this?

The claim originates in 1991 with a study conducted by Diane Halpern and Stanley Coren, published in The New England Journal of Medicine, "Handedness and Lifespan" and in a more rigorous form by the same authors in Psychological Bulletin. (Coren's previous article on the increased risk of accidents in left-handers was published in the American Journal of Public Health, in 1989.) These studies indeed make the claim that being left-handed correlates with decreased lifespan. The authors speculate that it might have something to do with the design of cars, for example, or knives – in other words, that left-handed people are the victims of products designed for right-handers.

The claim is legitimated by many sources. Besides being published in one of the premier medical science journals on the planet, it was consequently reported by Reuters and then printed by many leading news sources (see for example the NY Times). The statistical analysis was called into question very soon after the initial publication, but even decades later it seems that the refutation needs to be broadcast again and again. This indicates that despite the claim being false, it is surprisingly durable. (This motivates a further question: What makes people hold onto false claims and fail to grasp true claims?)

Here's a surprising claim that turns out to be true: Most published research is false. Read that again. The majority of published scientific research can't be replicated. In other words, if you hear about a study – coffee causes cancer, cellphones cause cancer, Google makes you stupid, whatever – there's a betting chance it isn't true. Surprising!

If you want to explore the reason for this, read the theory (Ioannidis, 2005a, "Why Most Published Research Findings Are False"). Ioannidis (2005b) shows that even for the highest-quality randomized trials, around 30% are eventually contradicted. Of course, lots of science isn't high quality or randomized – especially in education research. But even in a field like pharmaceutical research, where the financial stakes for truth are very high, a dismal number of "findings" can be replicated.

All this is to say that:

Ultimately, while surprise is a nice turbo boost to learning, we also need to help students understand that when we create new knowledge, it need not be surprising. Sometimes what we need to do is simply make steady, unremarkable, unsensational progress. As Jessica Love puts it:

The counterintuitive has its place. But our love affair comes at a cost. It leaves little room in the public consciousness for social scientific work that is incremental, for work that shores up and teases apart, for work that complicates, for work on the boundary conditions—those fragile social and mental habitats upon which decisions turn. In other words, it leaves little room for most of social science.

Surprise should always be leveraged as a way to increase self-reflection and critical thinking.