"Whenever you learn something new, your brain changes in a variety of ways. Until now we didn't know exactly how those changes were linked to behaviour," says study co-author Professor Allison Sekuler of U of T's department of psychology. Researchers have long known that people's ability to make perceptual judgements improves with practice, but until now they have never known why.

One theory suggests that learning decreases the variability in the way people's perceptual systems encode sights, smells, and sounds -- perception becomes more consistent and reliable. Another theory proposes that individuals become more efficient at encoding relevant information, and therefore become more sensitive to small differences among objects.

"As an analogy, imagine watching two televisions with poor reception," says co-author Professor Patrick Bennett, also of the department of psychology. "In one case, the programs from two or three stations are mixed up on one channel, so the image is visible but very unstable. In the other case, only one station is shown, but its signal is so faint that you can barely see it. The way to improve the first television is to increase its stability -- which means decreasing its variability. The way to improve the second television is to increase its sensitivity to the station's signal. The question is: Which kind of improvement produces perceptual learning?"

To answer this question, researchers trained people to identify photos of 10 unfamiliar human faces embedded in varying degrees of visual camouflage resembling the black and white static on an untuned TV. After several days, performance improved by up to 400 per cent and people were able to identify faces that were nearly invisible to them at the start of training. To see if this improvement was caused by a reduction in the brain's variability, the researchers presented the same sequence of patterns again and measured response consistency.

"If learning reduces the brain's variability, the response consistency would have increased with practice, but it didn't," Bennett says. The failure to see any change in response consistency demonstrates that learning occurred because the people were becoming more efficient at encoding the visual information relevant to face identification.

The same results were found in a second study that used abstract visual patterns rather than faces. "The fact that we get the same effect with two very different kinds of stimuli suggests that we've tapped into a really basic property of learning," says Jason Gold, co-author and graduate student in the psychology department.

"Now that we have this knowledge, the next step is to determine at what precise levels in the brain these processes occur," says Sekuler. "This fundamental knowledge may help us develop techniques to investigate and improve many other types of learning in people -- from learning to read to playing the violin."

This study was funded by grants from the Natural Sciences and Engineering Research Council of Canada.

CONTACT:

Megan Easton
U of T Public Affairs
(416) 978-5948
megan.easton@utoronto.ca

Note: If no author is given, the source is cited instead.

• Children's Health
• Brain Tumor

• Educational Psychology
• Intelligence

• Social cognition
• Psychophysics
• Learning disability
• Memory
• Attribution theory
• Developmental psychology