Two arrays of colored squares flash. Spot whether one changed. Measures visual working memory capacity (Cowan's K).
Change detection is the cleanest measure of visual working memory capacity (Luck & Vogel, 1997).
You see an array of colored squares for ~250ms, then a brief blank, then the same array (or with one color changed). Decide same or different.
Cowan's K formula estimates how many items you can hold: K = (hit rate − false alarm rate) × set size.
Most adults max out at K ≈ 3-4. Higher set sizes reveal your true capacity ceiling.
Luck & Vogel (1997, Nature) introduced the paradigm. It has become the most cited working memory capacity test in vision research.
Capacity correlates with prefrontal-parietal connectivity strength and is sensitive to attention disorders.
It isolates how much visual detail you can hold across a brief blank gap.
The blank interval between arrays is the key: it forces you to compare the second display against an internal memory of the first, not against an afterimage still on the screen. Accuracy reflects how many colour-location bindings survived that gap. Doing well means you stored several intact items, not just a vague gist of the array.
Unlike visual search or a simple matching task, nothing stays visible to scan, so you cannot offload the work onto the screen. This taps the same storage that lets you keep a few items in mind, such as remembering several icons after a menu closes or holding a short set of cards in a game between glances.
Push set size upward only once your accuracy at the current size is reliable.
Begin with small arrays of two or three squares and aim for consistently correct same-or-different judgements before adding a fourth or fifth. Try to take in the whole array as one snapshot during the brief display rather than naming colours one by one, since labelling is too slow for the exposure time. Level up only when errors stay low.
A common mistake is fixating on a single square, which guarantees a miss whenever the change happens elsewhere. Spread attention across the array and resist the urge to guess the moment the second display appears; give yourself the full gap to compare. Verbal rehearsal does not help here, so trust the visual impression instead.
Read your largest reliable set size, not any single lucky trial.
Your useful metric is the set size at which you stay clearly above chance, since a same-or-different choice is correct half the time by guessing alone. Holding three or four items accurately is typical for visual working memory; staying sharp at five or six is strong. Watch your own trend across sessions rather than chasing one high result.
This estimate shifts with state: fatigue, distraction, or a noisy environment can lower it on any given day, so a single dip is not a verdict. It reflects short-term visual storage for this kind of array, not your verbal memory, reasoning, or intelligence, and it should not be read as a fixed ceiling on what you can hold.