Firm convictions dominate news headlines these days, but because of a phenomenon called the hypercorrection effect, strongly held ideas that turn out to be factually incorrect are actually easier to amend . Brain imaging is now shedding light on how people change their minds during hypercorrection, potentially revealing the best ways for us to learn from our errors.

To understand hypercorrection, says cognitive psychologist Janet Metcalfe at Columbia University, "suppose I ask you, 'What is the capital of Canada ?' and you say 'Toronto. ' I say, 'How confident are you?' and you say, 'Very highly confident.' When I then tell you that actually the capital is Ottawa, you're very likely to remember it— not just a few minutes later but weeks later, and maybe for much longer, we think."

Scientists reason that in hypercorrection, after people discover that ideas they felt very sure about were not in fact correct, the surprise and embarrassment they feel makes them pay special attention to alternative responses about which they felt less confident . People then go on to take the corrected information to heart, learning from their errors.

"In contrast, if I asked you a question to which you gave a not-very-confident answer, like, perhaps, 'What color does amethyst turn when it is heated?' and you say, 'blue' with low confidence, when I tell you that it's actually yellow, you're not very likely to remember it," Metcalfe says.

Given this model , to learn more about what happens in the brain during hypercorrection, Metcalfe and her colleagues focused on brain regions linked to attention as well as those involved in metacognition (self- awareness of  the thought process ) . The researchers used functional magnetic resonance imaging (fMRI) to scan the brains of 14 volunteers while they answered nearly 600 general information questions that had single-word answers. The participants then rated their confidence on their responses.

"My favorite—'What is the last name of the Oscar award –winning actor who thanked his parents for not using birth control?' '[Dustin] Hoffman,'" Metcalfe says.

The scientists found evidence supporting their hypercorrection model. Both wrong answers and right answers lit up the anterior cingulate and medial frontal gyrus, parts of the brain linked with attention and metacognition .

"The anterior cingulate registers our surprise and maybe something that we might, roughly, call embarrassment, and so we gear up all our resources to better encode 'Ottawa,'" Metcalfe says, referring to her previous geography quiz . The region did not, however, activate as strongly for wrong answers about which subjects initially felt low confidence, suggesting that the participants would be less likely to remember corrections to such answers.

The medial frontal gyrus is involved in social processes, suggesting a role in hypercorrection is as well. "This makes a lot of sense—a lot of our knowledge comes from other people and books, and from consensus and encyclopedias, and Scientific American," Metcalfe says. "Even though in our experiments answers were delivered by a computer, those answers were written by people. So it makes total sense that accepting corrections involves your relationship with other people." Medial frontal gyrus activation patterns mirrored those of the anterior cingulate.

In addition, after people were told that an answer in which they were very confident was wrong, the fMRI showed activation in the right temporoparietal junction, an area linked with thinking about what others might know, and the right dorsolateral prefrontal cortex, a region linked with the avoidance of thinking about something. The former suggests that subjects recognized that others had different beliefs than them, wh ereas the latter hints they may have been suppressing their wrong answers after learning they were incorrect, Metcalfe says. The scientists detailed their findings online March 27 in the Journal of Cognitive Neuroscience.

The findings have implications for educational techniques and theory. "The broadest conclusion we might draw from these findings is that we may have the wrong attitude toward errors," says cognitive psychologist Robert Bjork at University of California, Los Angeles, who did not take part in this work. "Throughout society and our educational system, there tends to be an attitude that you don't want people making errors and mistakes during learning. These findings and related findings suggest that in order to increase the effectiveness of long-term learning and understanding, we should structure instruction and training so that likely errors and misconceptions will come up during the learning process, and use them as opportunities for learning."

He added, "when it comes to, say, job contexts such as nuclear power plants or the military or the police...we don't really want such errors to be deferred until a time and place where they may really matter, and matter greatly."

10 Comments

1. 1. JamesDavis 09:30 AM 4/27/12

   That may prove to be a good way of teaching, but what if you get the answer correct and the teacher gives you the wrong answer believing that their answer is correct? Will that confuse you and cause you to seek therapy when you turn 70? I know that the chance of that happening is slim, but the possibility is always there.

   Reply | Report Abuse | Link to this

2. 2. curmudgeon in reply to JamesDavis 10:41 AM 4/27/12

   Not so slim as you imagine! I remember with complete clarity the occasion when I was 10 years old on which my answer did not agree with the given answer but was in fact correct. It did not result in my need for therapy but I don't think it did much for the teacher's mental health!

   Reply | Report Abuse | Link to this

3. 3. curmudgeon 11:00 AM 4/27/12

   A fascinating study but I worry that it appears to say something quite contrary to actual life experience that correction often leads to further entrenchment and greater conviction in the 'wrong' answer. I think the problem is that the study is concerned only with simple matters of absolute fact which can be demonstrably shown to be false. These actually play a surprisingly small part in our education and world image. Our convictions tend to be based on a complex web of interacting beliefs and are frequently totally invulnerable to correction as a result. I would therefore worry that "likely errors and misconceptions" becoming part of the the learning process might lead to reinforcement rather than correction. Human beings are capable of believing the most extraordinary things in spite of overwhelming evidence to the contrary. Should we really put our faith in a system that encourages them to become convinced of the truth of errors in the hope of reaping the benefit of hypercorrection?

   Reply | Report Abuse | Link to this

4. 4. doctortdm 11:38 AM 4/27/12

   The article reminds me of a quote from Goethe "To guard from error is not the instructor’s duty, but to lead the erring pupil; nay, to let him quaff his error in deep, satiating draughts, this is the instructor’s wisdom. He who only tastes his error, will long dwell with it, will take delight in it as in a singular felicity; while he who drains it to the dregs will, if he be not crazy, find it out."

   Reply | Report Abuse | Link to this

5. 5. nmlevesque 01:59 PM 4/27/12

   "In addition, after people were told that an answer in which they were very confident was wrong, the fMRI showed activation in the right temporoparietal junction, an area linked with thinking about what others might know, and the right dorsolateral prefrontal cortex, a region linked with the avoidance of thinking about something. The former suggests that subjects recognized that others had different beliefs than them, whereas the latter hints they may have been suppressing their wrong answers after learning they were incorrect, Metcalfe says."
   
   Or they are trying to avoid thinking that other people's different views are correct. How would you tell the difference between these two possibilities?

   Reply | Report Abuse | Link to this

6. 6. jtdwyer 01:03 AM 4/28/12

   I wonder how these researchers would have responded if asked whether their selected subject population adequately represented any general population...
   
   Once again (likely resulting from the cost of testing), general conclusions about the human population are derived from fMRI tests of 14 (unpaid) "volunteers" (likely young university students).
   
   The article described an unusual variation of the scientific method:
   "Scientists reason that in hypercorrection, after people discover that ideas they felt very sure about were not in fact correct, the surprise and embarrassment they feel makes them pay special attention to alternative responses about which they felt less confident."
   
   I reason that this model best applies to students trained to fear embarrassing correction by authority figures in a classroom of peers...
   
   Dedicated young students might be predisposed to abandoning any beliefs identified by professors as being incorrect, or at least inconsistent with those expected by the university. Test results obtained from a subject population consisting of experienced engineers or Navy Seals, for example, might have been different.

   Reply | Report Abuse | Link to this

7. 7. sct999221 10:30 AM 4/28/12

   And then there's this:
   
   "Facts don’t necessarily have the power to change our minds. In fact, quite the opposite. In a series of studies in 2005 and 2006, researchers at the University of Michigan found that when misinformed people, particularly political partisans, were exposed to corrected facts in news stories, they rarely changed their minds. In fact, they often became even more strongly set in their beliefs. Facts, they found, were not curing misinformation. Like an underpowered antibiotic, facts could actually make misinformation even stronger."
   
   https://mail.google.com/mail/u/0/#inbox

   Reply | Report Abuse | Link to this

8. 8. RockyBob 11:49 AM 4/28/12

   I'd agree that "important" erroneous beliefs might become more entrenched, and that in order to change such beliefs it is necessary to create major dissonance. In the case of astrological beliefs, I'm reminded of the Oregon professor who provided "individualized life and future assessments" for each of his students based on each students birthdate, city of birth, etc. Only after the students had been thoroughly dazzled by the accuracy of his assessment (average accuracy rated around 90%, as I recall) for each individual did the students then find that all the assessments were all identical. When confronted with real, immediate dissonance, apparently many of the students revised their beliefs in astrology.

   Reply | Report Abuse | Link to this

9. 9. julianpenrod 08:17 PM 4/29/12

   On Christmas Day, 2011, a house in Stamford, Connecticut bvelonging to ad executive Madonna Badger burned to the ground, killing her parents and children. Authorities refused to allow a forensic examination of the house and, instead, had it razed to the ground. Later, they claimed live embers in a bag of fireplace debris left in a mudroom caused the fire. They never saw any live embers in the bag, they never saw the bag! They simply supplied that as the "official story" for the event and left it to the dim to accept it as absolutely proven. Not once did any who hold themselves forward as defenders of "science", which they hold forward as the only reliable way to find the truth, emphasizing such things as proof and "reason", take exception to this patent example of reqaching "conclusions" without proof. "Science' isn't about truth, it's about saying what the New World Order wants the dullards to believe. Trusting "science" may be among the most dangerous things someone can do.

   Reply | Report Abuse | Link to this

10. 10. mascampoamor 06:24 AM 5/1/12

    indeed, feeling is the best experience and experience is the best teacher.

    Reply | Report Abuse | Link to this