A grid of items appears. One is the target (e.g., a green circle among red circles and green squares). Find it as fast as you can. Half the trials are "feature search" (target pops out), half are "conjunction search" (target requires serial attention).
What makes this one different
We sample set sizes 8, 16, 24 and fit your RT × set-size slope. Feature search should give a near-flat slope (target pops out). Conjunction search gives a steeper slope. The two slopes compared is the cleanest measure of attentional capacity.
Press 1 if target present, 0 if absent. Or click the buttons.
RT slope across set size 8, 16, 24. Feature search should be near-flat; conjunction search is steeper.
No sessions yet.
The classic Treisman feature-integration theory predicts that searching for a unique feature (e.g., a red item among green items) is essentially parallel and "pops out," while searching for a conjunction of features (e.g., a red X among red O's and blue X's) requires serial attention. We let you see that difference in your own data.
Feature-search slope near 0 ms/item is expected. Anything above ~5 ms/item suggests you're not finding the pop-out.
Conjunction-search slope of 15–30 ms/item is typical. Above 40 ms/item suggests serial search is slow.
Catch trials. If you just say "yes" to every grid, your accuracy is meaningless. About 1 in 6 trials has no target.
Either the feature/conjunction distinction wasn't strong enough for you, or you're searching serially even when you don't need to. Both are real patterns.