Task Switching Test: How Much Does Changing Rules Cost You?
Cued task switching is the standard way to measure cognitive flexibility. A colored shape appears with a word above it telling you which rule applies: answer by color, or answer by shape. The same shape needs a different answer depending on the rule, and the rule changes without warning. This test separates two different costs that most versions blur together — what it costs to change rules on this trial, and what it costs merely to be holding two rules at once. Free to play, scored instantly, no sign-up required.
A quiet round of Task Switching Test — Standard. When you start, everything else fades — just you and the task.
No sign-up needed. Press Esc to exit anytime.
Tip
Read the word, not the shape, first.. The shape arrives 300 ms after the word. That gap is there on purpose: it is time to load the rule before the stimulus lands. People who spend it get a much smaller switch cost than people who wait for the shape and then look up.
Go Pro — train without limits
- ·Unlimited sessions
- ·Every difficulty level
- ·Full growth history
- ·Ad-free training
How to play
- A word appears first: COLOR or HUE means answer by color; SHAPE or FORM means answer by shape. The two words in each pair mean the same rule.
- A colored shape follows. Press ← for teal or circle, → for amber or square — whichever the current rule asks for.
- The first two blocks use one rule each. The last block mixes them, and the rule changes on about one trial in four.
- Finish all trials to see your accuracy, your speed, and the two switching costs.
Task Switching Test tips
Ways to play it better — each one comes from how this version actually scores, not from general advice about concentration.
Read the word, not the shape, first.
The shape arrives 300 ms after the word. That gap is there on purpose: it is time to load the rule before the stimulus lands. People who spend it get a much smaller switch cost than people who wait for the shape and then look up.
Two words, one rule — that is deliberate.
COLOR and HUE are the same instruction, and so are SHAPE and FORM. Having two words per rule is what lets the test tell apart 'you changed rules' from 'you read a new word'. Without it, the time spent reading an unfamiliar cue gets counted as the cost of switching, which inflates it — sometimes to the point where the real cost is not there at all.
The mixed block is slower even when nothing changes.
Compare a repeat trial in the mixed block with a trial from a pure block: the mixed one is slower even though the rule did not change. That gap is the mixing cost — the price of keeping two rules available at once, separate from the price of actually swapping.
One caveat worth knowing: a change of strategy moves your score for reasons that have nothing to do with attention. Your dashboard reads each session against your own recent ones, so if you adopt something from this list, give it a few sessions before reading anything into the trend.
Frequently asked questions
What is the difference between switch cost and mixing cost?+
Switch cost is local: how much slower you are on a trial where the rule just changed, compared with a trial where it did not. Mixing cost is global: how much slower you are in the mixed block overall, compared with a block that only ever used one rule — even on trials where nothing changed. They come apart because they reflect different things: one is the act of swapping, the other is the load of holding both. Most online versions report only a single blended number.
Why does the rule change on only a quarter of the trials?+
Because switching has to be the exception for it to cost anything. When the rule changes half the time, people stop settling into one rule and simply hold both ready, and the measured cost shrinks toward nothing — this was shown directly in the research literature. Keeping changes rare is what makes the default 'stay on this rule', so that a change is a real event with a real price.
How is this different from the Trail Making Test Part B?+
Part B — alternating between numbers and letters — is also a task switching task, and it is a good one. But it has no single-rule block to compare against and no trials where you stay on the same rule, so it can only give you one blended number. This test is built to pull the two costs apart. If you want the pen-and-paper classic, Part B is on this site too.
Is a big switch cost bad?+
Not in any way you should read into. Switch and mixing costs are difference scores — one average subtracted from another — and difference scores are notoriously unstable for a single person on a single day. Yours can move a lot between sessions without anything about you changing. That is exactly why they are shown here as information and are deliberately left out of your score and out of the leaderboard.
Is this a diagnostic test?+
No. Nothing on this page is a screening instrument. It is a practice and self-tracking exercise: not validated, not normed, and it cannot screen for or indicate any condition. The trial counts here are far shorter than a research version, which is a deliberate trade-off for a three-minute round, and it means your numbers cannot be compared with published figures.
Where this fits
- Processing Speed Test — Switching on its own, with the scanning taken out. Part B blends the cost of changing rules with the cost of holding two of them ready at once; this one has a single-rule block to compare against, so it can report those two separately.
Try another challenge
Schulte Table
Scan a scrambled number grid as fast as you can, in order. Practices and measures visual search speed and peripheral vision.
Stroop Test
Name the ink color, not the word. Practices and measures selective attention and mental control.
Visual Search
Find the one target hidden among distractors. Practices and measures scanning speed and sustained attention.
Multiple Object Tracking
Keep several identical moving dots in mind at once. Practices and measures divided and sustained visual attention.
Simon Effect Test
Answer by color, ignore which side it appeared on. Practices and measures response conflict and inhibitory control.
Flanker Test
Read the middle arrow, ignore the four crowding it. Practices and measures selective attention and interference control.
Go/No-Go Task
Tap for circles. Hold still for squares — that's the hard part. Practices and measures inhibitory control and impulse braking.
Sustained Attention to Response Task
Tap every digit, hold still for one. Ten minutes of focus in two. Practices and measures sustained attention and mind-wandering resistance.
Mackworth Clock Test
A hand ticks round a blank dial. Catch the moment it skips. Practices and measures vigilance and rare-event detection over time.
Psychomotor Vigilance Task
Stop the timer the moment it appears. Again, for three minutes. Practices and measures alertness and how well it holds up.
Cancellation Task
Sweep a dense field and cross out every target. Practices and measures visual scanning and selective attention.
Trail Making Test
Connect the sequence — Part B alternates numbers and letters. Practices and measures processing speed and attention switching.
Change Blindness Test
Two versions flicker back and forth — spot what changed. Practices and measures attention allocation across a scene.