Cognitive Psychology
Change Blindness
Two versions of the same scene, alternating. With a blank between them, a very large difference can take a surprisingly long time to find. Take the blank away and it leaps out.
Original vector scenes — a demonstration, not a measurement of anybody
Learning objective
By the end you should be able to explain what the blank interval does to change detection, and why that does not mean people store nothing about a scene.
About 20 minutes. Nothing you do here is saved or sent anywhere.
Flicker, safety and pacing
The alternation is slow and can be switched off. A full cycle takes about 1.7 seconds, so the screen changes well below the three-per-second threshold associated with photosensitive reactions. There is a Pause button throughout, and a Step manually mode in which nothing changes until you press a key or a button.
If your system asks for reduced motion, this starts in manual mode and says so. The demonstration works perfectly well that way.
- Answer the prediction below — it unlocks the scenes.
- Pick a scene, start it, and search for the one thing that differs between the two versions.
- Press I can see it as soon as you have it, then say what changed.
- Run the same scene again with the blank switched off, and notice how long that takes.
- Each scene works properly only once. Four are provided, and used ones are marked.
First, a prediction
You are about to see two versions of a simple street scene, alternating with a blank grey screen between them. Exactly one thing differs. Some of these changes are large, and obvious once you have found them. People still miss them while the picture keeps flickering.
The flicker task
Find the one thing that changes
Two versions of a scene alternate. Search for the difference, then say what it was.
Status Choose a scene and press Start.
Answer the question above to unlock the scenes.
Your trials so far
Every trial you have run
| Scene | Change | Where | Size | Blank? | Time | Identified? |
|---|
Challenge — why does the blank matter?
What this demonstrates
The transient is doing the work, not the picture
When something in view changes, the change itself produces a local signal at the place where it happened — a flash, a flicker, a movement. That signal pulls attention there without any searching. A blank screen wipes that advantage out by producing the same signal everywhere at once. The picture is identical either way. What changed is whether the change announces its own location.
Comparison needs attention
With the local transient gone, finding the difference becomes a deliberate search: hold part of the scene in mind, look again, and compare. That is slow, effortful and easy to do in the wrong place. Where the change is, how large it is, and whether it is somewhere you would naturally look all matter — which is why the trials in this tool differ on exactly those properties.
This does not show that nothing is stored
The strongest reading of change blindness — that people retain essentially nothing between glances — does not follow, and the evidence is against it. People remember a good deal about scenes they have looked at, including things they failed to report a change in. What the paradigm shows is that comparing two views requires attention to be in the right place at the right time; not that the views leave no trace.
Failing to report is not failing to see
As with the other awareness paradigms, the measurement is a report. Someone may hold information they cannot bring to a deliberate comparison, or may compare the wrong region confidently. A search that takes forty seconds is not forty seconds of blindness — it is forty seconds of looking in other places.
For teaching elsewhere: take this activity as one self-contained block of HTML, on the clipboard or as a file. Either way it is styled so that it will not disturb the page you put it into.