Change blindness: why people miss what changed

In 1997 Ronald Rensink, Kevin O’Regan and James Clark showed people two photographs, alternating, with a blank screen of 80 milliseconds between them.

Without the blank, people spotted the difference in under a second, because the change shows up as movement and pulls the eye straight to it. With the blank, the movement is gone, and people searched for eleven seconds on average¹, sometimes close to a minute, for changes that were obvious once seen. Changes to the things people found interesting were found faster than changes to the background.

The lesson is that we do not keep a detailed picture of the world in our heads to compare against. We see what we are attending to, and we notice change when something moves.

In 1998 Daniel Simons and Daniel Levin had a stranger stop people on a university campus to ask for directions.

Partway through, two men carrying a door walked between them, and the stranger was swapped for a different person. Only 7 of 15² people noticed. Those who did were students of about the experimenters’ own age; the people who missed it were older. When the experimenters dressed as construction workers and approached students, only 4 of 12 noticed. We look more closely at people we see as like ourselves.

Expertise does not protect us. In 2013, researchers hid a gorilla, 48 times the size of a typical nodule, in a lung scan. 20 of 24 radiologists⁴ did not report it, and eye tracking showed that 12 of them had looked right at it. That is a related effect, inattentional blindness: missing something unexpected while concentrating on a task.

A web page does to its readers what the flicker test does on purpose.

A full reload blanks the screen, and when it returns a price has changed, a total has gone up or an error has appeared somewhere else. An item added to a basket updates a small number in a far corner. A warning sits in a box that looks like an advertisement, the place people have learned not to look, which researchers named banner blindness⁵ in 1998.

People who use screen readers miss these changes entirely unless the page announces them, which is why the accessibility guidelines require status messages⁷ that assistive technology can read out without moving the person away from what they were doing.

This one runs here, on this page: it measures you, not a design. Use what it shows you when you review a prototype or a live page with the other tools.