Selective attention, with examples from daily life and from games
· 6 min read · Tyamka team
What selective attention is, the classic experiments behind it, and concrete examples, from a noisy café to the Stroop test, odd one out and card sorting games.
Selective attention is picking out what matters for the task at hand and letting the rest fade. You follow one voice in a noisy room, spot your bag on a crowded luggage belt, read a page while the TV is on. The information you ignore still reaches your eyes and ears, and researchers still argue about how much of it gets processed. A simple way to watch it work is a search game such as odd one out game online, where one figure in a grid is different and everything else is noise.
Examples from daily life
- A conversation in a loud café. You hear the person across the table and not the dozen voices around you. Yet if someone at the next table says your name, you may well notice.
- Looking for your keys. You scan a cluttered table for one shape and one color, and the other objects barely register.
- Reading with the radio on. The words on the page get your attention, the song becomes background, until a line you know pulls you away.
- Driving. You watch the lane, the car ahead and the signals, and ignore the adverts on the roadside, mostly.
- A form with a mistake. You check a column of figures for the one that looks wrong, and the correct ones blur together.
Each example has the same structure. There is a target, there are distractors, and you steer your attention toward the target, by a feature, by a place, or by a rule.
The classic experiments
The cocktail party problem. In 1953 the British scientist Colin Cherry played two different messages to people through headphones, one to each ear, and asked them to repeat one message aloud as they heard it. They could do it, and afterwards they could say almost nothing about the content of the other message, and did not notice when it switched from English to German. They did notice if the voice in the ignored ear changed from a man to a woman. In 1959 Neville Moray found that some people, about a third in his study, picked up their own name from the ignored ear.
Filters and attenuation. Donald Broadbent proposed in 1958 that attention works as a filter that lets one channel through and blocks the rest. Anne Treisman argued in the 1960s that the ignored channel is turned down rather than switched off, which explains why your name can still get through.
Features and conjunctions. In 1980 Anne Treisman and Garry Gelade described two kinds of visual search. Finding a red item among green ones is quick, and it barely matters how many green items there are, because the red one stands out. Finding a red circle among red squares and green circles is slower, and the time grows with every item added, because no single feature marks the target. Their explanation was that such search needs focused attention on items one at a time, or close to it.
What you do not see. In 1999 Daniel Simons and Christopher Chabris asked people to count passes in a video of a basketball game. About half of them missed a person in a gorilla suit walking through the scene. Attention on one task can leave a large, obvious event unseen.
Three ways to select
From these experiments it helps to sort selective attention into three rough kinds.
| Selecting by | What it looks like | Example |
|---|---|---|
| Feature | you look for a color, a shape, a size | the red car in a car park |
| Place | you watch one spot and ignore the edges | the middle arrow in a row of arrows |
| Rule | the task tells you which feature counts right now | name the ink, not the word |
Real tasks mix them. The games below each lean on one more than the others.
Examples from the games on this site
Odd Shape, finding the one that differs
In our odd one out game, called Odd Shape in the app, one figure in a grid is different and you tap it before the round's time runs out. On levels 1 to 10 all the other figures are the same. The odd one can be turned, larger or smaller, lighter, mirrored, or drawn with one or two pixels changed. The grid grows from 3×3 to 6×6 and the difference gets smaller.
Levels 11 to 15 change the search and start again from a 4×4 grid, growing to 7×7. The other figures come in two groups, and the odd one shares one feature with each group, so no single feature gives it away. This is the conjunction search Treisman and Gelade described. We designed these levels so that bigger grids take longer to search, which is still to be checked in playtests.
Stroop test, ignoring part of one object
In the Stroop test the distractor is not beside the target, it is the target. You name the ink of a color word and hold back the reading of the word. This is selection by rule inside one object, and it is described in the Stroop effect explained.
Arrow Rule, a rule against a habit
In Arrow Rule one arrow appears with a rule, Same or Opposite, and you answer by the rule. On the last levels the arrow sits off center, and its position pulls you toward the wrong side. How this differs from the classic row of flanker arrows is in the flanker task and the arrow Stroop.
Two Piles, sorting by the rule of the moment
In our card sorting game online, Two Piles in the app, cards come one at a time and you send each one left or right. By shape, circles go left and squares go right. By color, yellow goes left and blue goes right. Yellow cards are solid and blue cards are hollow, so the color rule can also be followed by fill. Blue circles and yellow squares are the tricky cards, because shape and color send them to opposite sides. From level 8 the rule above the card switches between cards, three times per level at first and twelve times on level 15, so you have to check which feature counts right now.
This sounds like the Wisconsin Card Sorting Test, which Esta Berg and David Grant described in 1948, building on an older sorting test by Egon Weigl, and there is one big difference. In the Wisconsin test you have to work out the rule from feedback. In Two Piles the rule is always written above the card. The game is about applying a visible rule quickly, not about discovering a hidden one.
Schulte table, searching in order
A Schulte table is also a search task. You look for the next number among many others, and every number you have already found becomes a distractor. See what a Schulte table is for the details.
What the games do and do not say
Each of these is a short timed game built on one narrow attention rule, and your points show only how you did on that level on that day. They move with sleep, practice, time of day and the device you play on. We do not claim that playing them improves attention outside the game. What research says on that question is in does brain training work. Treat your score as a record to beat against yourself, not as a measure of your attention.
If you want to see selective attention in action, start with a few rounds of odd one out and notice how your search changes on the late levels. All the other games are on the all games page.