Change Blindness Illusion

A flicker that hides the obvious

In the demo above, two photos of the same living room alternate, with a brief grey blank between them. They're identical except for one object, and it isn't a subtle difference: it's a sizeable object in plain view. In the original experiments the changes were just as blatant, such as a jet engine disappearing from an aircraft's wing.

Most people still take a surprisingly long time to find it, sometimes many seconds, sometimes not until it's pointed out. Once seen, it's impossible to miss, which makes the earlier failure seem baffling. That gap between how obvious a change is once found and how invisible it was a moment before is called change blindness.

You hold less of the scene than you think

Change blindness isn't a problem with the eyes. Sharpness, contrast and motion detection are all working normally. The limit is in how attention and short-term visual memory build what feels like a rich, complete picture of the world.

That feeling is itself something of an illusion. At any moment you keep only a small part of a scene in enough detail to compare with what you see next, roughly whatever you're attending to. The rest is held as a rough summary: this area is a bookshelf, that one is a window. Normally that's enough, because the world doesn't change behind your back, and you can always look again to check. And when something does change while you're looking, the movement itself catches your attention automatically.

The flicker method removes that safety net. The grey blank hides the burst of motion that would normally announce the change, so the only way to find it is to compare your patchy memory of one picture with the next, area by area. That's slow and effortful, and it only works when you happen to be attending to the part that changed.

Two classic experiments

Ronald Rensink, Kevin O'Regan and James Clark introduced the flicker paradigm in 1997: an original photo and an altered version alternate, separated by a brief blank, and viewers press a button when they spot the difference. Large, meaningful changes often took many seconds to find. Without the blank, the same changes were found almost at once, which showed that hiding the motion signal was what made them hard (Rensink, O'Regan & Clark, 1997). The demo's checkbox lets you compare both versions.

Daniel Simons and Daniel Levin took the effect out of the lab. In their 1998 "door study", an experimenter stopped a pedestrian to ask for directions. Mid-conversation, two people carried a door between them, and behind it the experimenter swapped places with a different person, with different clothes and a different voice. About half the pedestrians carried on giving directions without noticing they were talking to someone else (Simons & Levin, 1998). The door did the same job as the blank screen.

Why it matters

Film editors rely on change blindness: continuity errors between shots, such as a glass that refills itself, mostly go unnoticed until someone freezes the frames side by side. The same limits help explain why witnesses can be confident about details of a scene that changed without their noticing. The closely related inattentional blindness goes a step further and shows that things in plain view can be missed even when nothing is hidden at all.