Cognitive Psychology
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
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.
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.
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.
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.
| Scene | Change | Where | Size | Blank? | Time | Identified? |
|---|
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.
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.
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.
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.