Task Switching
Alternate number/letter rules
About this trainer
Task Switching shows you a stream of items and keeps changing the rule you must apply. One moment you sort a number by odd or even, the next a cue flips you to sorting it by high or low, so on every trial you read the cue, drop the old rule, load the new one, and answer fast without losing accuracy.
What it develops
It trains cognitive flexibility and the executive control that lets you reconfigure a mental set on demand, the same machinery you lean on whenever you juggle two activities and have to keep dropping one to pick up the other.
History
The effect was first reported by Arthur Jersild in 1927, who timed people running through a mixed list of tasks versus a single repeated one and saw the mixed list cost more. The idea sat largely dormant until the mid-1990s, when a wave of cognitive-psychology work revived it as a clean window into executive control.
Who created it — and when
There is no single inventor of the exercise. Arthur Jersild introduced the comparison in 1927, and the modern version was shaped by Robert Rogers and Stephen Monsell in 1995, who coined the term switch cost and built the widely copied alternating-runs paradigm.
How to train
Read the cue before you look at the item, and name the current rule to yourself so the wrong rule does not leak in. Push your speed only as far as your accuracy holds, since rushing inflates errors rather than reducing the switch cost, and lengthen the gap before each switch when you want to practise preparing in advance.
How long to practise
Short focused blocks work best, roughly 5 to 10 minutes at a time, a few days a week. Switching is mentally taxing, so quality drops quickly once you fatigue, and stopping while still sharp beats grinding a long session.
Evidence base
What is solidly shown is exactly what you would expect: the switch cost is a robust, reliable effect, and with practice you get faster and more accurate at the switching task itself. The bigger promises are far shakier, as broad transfer to everyday multitasking, general intelligence or protection against age-related decline is weak and contested, and the most careful reviews of brain training, such as Simons and colleagues in 2016, conclude that far-transfer claims are largely unsupported, so treat them with caution.
Recommendations
Treat the cue as the real task: pause a beat to register the rule before answering, and your error rate will fall faster than your reaction time.
FAQ
Why am I slower right after the rule changes?
That slowdown is the switch cost, the time your brain needs to drop the old rule and load the new one. It is normal and is exactly what the exercise measures and trains.
Will this make me better at multitasking in real life?
It reliably makes you better at this task, but evidence that the skill carries over to everyday multitasking is weak and disputed. Train it because the practice is sharp and engaging, not on a promise of broad real-world transfer.
Should I go for maximum speed?
No. Push speed only as far as your accuracy holds. Rushing mostly adds errors, while a brief read of the cue before you answer is what actually shrinks the switch cost.
Variants
Versions differ by how the switch is signalled. Alternating runs follow a fixed pattern like AABB so you can predict the switch, explicit cueing shows a symbol each trial, and voluntary switching lets you choose the rule yourself. Difficulty also scales with the number of rules and the time you get to prepare.