Doug Crawford’s team sees through ‘change blindness’


(Above: lab where Prof. Doug Crawford and his research team conduct experiments)

There are times when we are so preoccupied that we don’t hear a friend speak to us, or we don’t take note what someone is wearing. But what does it mean when major changes happen around us – and we don’t notice them at all?

Known as “change blindness”, this phenomenon has been the subject of study by Prof. Doug Crawford (left), York’s Canada Research Chair in Visuomotor Neuroscience and faculty member in York’s Centre for Vision Research (CVR). His research team included Matthias Neimeier (below, left), a post-doctoral fellow from Tuebingen, Germany, and U of T physiology Professor Douglas Tweed (below, right), who is also an adjunct member of CVR. Tweed, senior author of the article, provided the theoretical expertise for the project.

Recently, Crawford, Neimeier and Tweed had an article on their research printed in the March 6 edition of the prestigious, scientific journal, Nature.

The article is entitled “Optimal transsaccadic integration explains distorted spatial perception”. Their findings were based on experiments carried out in one of Crawford’s CVR labs. The research project was funded by the CIHR Group for Action and Perception, to which Crawford and Tweed belong.

Crawford submitted the following information about the project to YFile.

In recent years, there has been considerable excitement in the vision and cognition communities about “change blindness”. This is a phenomenon where people fail to observe changes in the environment when there is some kind of visual distraction during the change.

In one dramatic study, unsuspecting subjects who were giving directions to an experimenter failed to notice when this experimenter surreptitiously “changed” into another person, even to another sex, after the visual interruption of two people carrying a door between the subject and experimenter.

This, of course, is also an important principle in sleight-of-hand magic tricks. But psychologists and neuroscientists have interpreted this to mean that, despite intuition, the brain does not retain a very good moment-to-moment memory of the visual world.

We reasoned that this is not a case of bad memory, but rather a case of a very sensible brain rejecting possibilities that are extremely unlikely to occur (such as people switching sex mid-way through a conversation).

To test this we looked at the simple case where people fail to notice small, quick jumps in a visual stimulus when they occur during an eye movement. First, we trained mathematical models of the brain to remember where objects are. These models predicted that when a stimulus jumps during an eye movement, the brain would discount the apparent change as a mistake in the eye movement rather than a less likely event in the external world.

Specifically, our models predicted that subjects with more accurate eye movements would be more sensitive to perceiving stimulus jumps than subjects with more sloppy eye movements.

Detailed testing in real subjects showed this to be true. This casts doubt on the idea that change blindness is necessarily due to poor memory; at least some forms are due to the brain making optimal inferences about the visual world.