What is a visual search test?
In a visual search test you look for one target among distractors and respond as soon as you find it. Psychologists use tasks like this to study how attention moves around what we see. Here the target is a single letter T, and the distractors are Ls turned every which way, so the T can't be spotted at a glance and has to be searched for.
How this test works
There are 20 searches. Each one shows 6, 12, 18 or 24 shapes, and each of those numbers comes up 5 times, in random order. Between searches a small cross holds the center of the box for 0.7 to 1.2 seconds. When the shapes appear, find the T and tap or click it. Tapping an L ends that search as a miss and moves on.
How is your result calculated?
Each search is timed from the animation frame in which the shapes are drawn to the moment your finger or mouse button goes down on the T, using the timestamp your browser gives the press. Your result is the average time of the searches where you found the T; misses are counted separately, in your number found out of 20.
You also see each extra shape: how much longer your searches took, on average, for every shape added to the screen. It is the slope of the straight line that best fits your own searches, plotted as number of shapes against time. It is calculated from your run only, and it is the number researchers look at in visual search: a slope near zero means the target stood out on its own, and a steep one means you had to look at the shapes one by one.
Why is the T so hard to find?
Because a T and an L are made of the same two bars. Only the way the bars meet is different, so nothing about the T's color, size or tilt gives it away. A red shape among green ones seems to jump out however many shapes there are. A T among Ls usually has to be checked shape by shape, which is why your searches take longer as the screen fills up.
The psychologists Anne Treisman and Garry Gelade explained this difference in their feature integration theory, published in 1980. In their account, simple features such as color are picked up across the whole scene at once, but combining features into one object needs attention to focus on each item in turn.
What can change your result
- Pointing. Once you have found the T you still have to reach it. A finger on a small phone screen travels less than a mouse across a big desktop screen, and both add time that isn't searching.
- Your search strategy. Scanning in rows usually works better than letting your eyes jump around, which tends to revisit shapes you have already checked.
- Screen size and distance. The shapes scale with your screen, and how much of the box you can take in at once changes with how far away you sit.
- Luck. If the T happens to sit where you look first, that search is quick. Averaging over 20 searches evens this out.
What this test does not measure
It measures how quickly you found one T among Ls on your device, in one sitting. It isn't a clinical attention test and can't diagnose anything, and results from different devices aren't directly comparable. For another test of attention, try the Stroop Test, or follow moving targets in Multiple Object Tracking.