Students Who Are Confidently Wrong
Jonathan Firth's Memory & Metacognition Updates #93
Hello! November already. I hope your term/semester is going well.
Last time I shared a post on ‘errorless learning’. Learning that proceeds in small steps and minimises errors is often assumed to be the best approach, but researchers such as Janet Metcalfe have argued that avoiding errors is actually unnecessary and can even be counterproductive (Metcalfe, 2017).
Today, I want to focus on a specific phenomenon mentioned in the research paper I shared last week – the hypercorrection effect.
We might assume that when students are wrong about something but very confident in their response, it would be hard to correct them. In fact, a range of research studies have shown that confident errors are easier to correct (Corral & Carpenter, 2022).
About the hypercorrection effect
Why on earth would a high-confidence error be easier to correct than a tentative, uncertain error? Generally we are confident about the things that we know well, and so we might assume that confident knowledge would be hard to change.
One possibility is that learners form an episodic memory of an error, as I mentioned last time. We can all remember a time when we messed up! The more foolish and blatant a mistake, the easier it is to recall. It grabs our attention.
Another possibility is that the student draws on related knowledge. As Metcalfe explains in her paper, participants often state that the right answer would have been their second guess! In contrast to tentative errors in areas where the learner is a novice, it may be easier for them to see where and why they went wrong.
The hypercorrection effect and metacognition
The hypercorrection effect reminds us that every time learners get things right or wrong, there is a metacognitive belief alongside. How confident are they in their answer?
Actually, this is something that metacognitive researchers often study. For example, we might say, “Who wrote Treasure Island?” and then ask, “How sure are you about your answer, on a scale of 1–10?”
This basic methodology of asking for student confidence is not inherently difficult to do, and could be adapted to any action research project.
As we have seen when discussing study skills and other matters in these updates, people’s metacognitive beliefs are not always accurate—learners can be confidently wrong about things.
Implications of the hypercorrection effect
The hypercorrection effect suggests that entrenched misconceptions about science, history or other areas need not be as hard to overcome as we would assume.
Certainly, misconceptions about a concept can affect how someone thinks about subsequent information. A cute example of this reported by Bransford et al. (2000, p. 10) concerns children being told that the earth is round rather than flat:
“Consider the challenge of working with children who believe that the earth is flat and attempting to help them understand that it is spherical. When told it is round, children picture the earth as a pancake rather than as a sphere... If they are then told that it is round like a sphere, they interpret the new information about a spherical earth within their flat-earth view by picturing a pancake-like flat surface inside or on top of a sphere, with humans standing on top of the pancake. The children’s construction of their new understandings has been guided by a model of the earth that helped them explain how they could stand or walk upon its surface, and a spherical earth did not fit their mental model.”
As new information can be interpreted in terms of (confident) misconceptions, it’s important to correct these. The hypercorrection effect is certainly relevant here! However, as Corral & Carpenter (2022, p. 209) warn in a report on three recent experiments, “the educational applications of this phenomenon have been largely unexplored”.
They go on to explain that what we know about the effect is mostly based on short-term lab studies with simple material. Accordingly, some questions where educators need clearer answers include:
Does this work for realistic curriculum materials?
Do the corrected responses persist in the long term?
Have learners really changed their mind, or just memorised the correct answer?
Corral and Carpenter’s experiments sought to tackle these issues. And there is some good news—they did find the effect with real curriculum material (from a psychology undergraduate course), it persisted after many weeks, and corrections did transfer to novel questions (meaning that students had really learned, not just memorised the answers).
However, although around 60% of corrections persisted over 10–15 weeks, in other cases students reverted to the original error (the researchers called these ‘return errors’), and this was more likely to happen with confident errors.
Therefore, students being overconfident but wrong was still an issue—something that may sound familiar to those of us who have graded papers across an academic year!
Here is the paper mentioned above:
Corral, D., & Carpenter, S. K. (2023). Long-term hypercorrection, return errors, and the transfer of learning in the classroom. Journal of Applied Research in Memory and Cognition, 12(2), 208–229.
By the way, this feels like an area where more research is needed in schools, and across a range of curriculum areas. If you work in the school sector and want to get engaged with research, it could be a great area for a classroom research project.
Have a great week,
Jonathan
Please note that my slides and similar materials are used under a CC BY-NC 4.0 license. This means you can use or adapt them with attribution for non-commercial purposes. If you wish to use my materials for other purposes, feel free to get in touch.