Task Switching

On each trial you see an item and a cue telling you which question to answer. Sometimes the cue asks about color; sometimes it asks about shape. The rules never change, but which rule applies keeps flipping. Your job is to answer the right question quickly, even when it just switched from the last one.

That flipping is the whole point. It measures cognitive flexibility: how smoothly you can drop one set of rules and pick up another. Switching is not free, and the size of that cost is exactly what the test captures.

• Rule holding: keeping both possible tasks ready in mind.
• Set shifting: letting go of the last rule and applying the new one.
• Interference control: ignoring the answer you would have given under the other rule.

The idea is old. In 1927, Arthur Jersild compared blocks of trials where people repeated one task against blocks where they alternated between two, and found that alternating was slower. That slowdown was the first clear sign that switching itself takes time.

The modern version comes from Robert Rogers and Stephen Monsell in 1995. They mixed switch and repeat trials within the same run and cued each one, which let them measure the cost precisely. The key number is the switch cost: your reaction time on switch trials minus your reaction time on repeat trials. If you answer in 700 milliseconds after a switch but 550 after a repeat, your switch cost is about 150 milliseconds. That gap is the price your mind pays to change rules.

The headline number here is not raw speed but the switch cost. A smaller cost means you change tasks with less friction. Almost everyone shows a positive cost, because switching genuinely takes extra time, but the size varies from person to person and from run to run.

• Small switch cost: you recover quickly after the cue changes.
• Larger switch cost: the flip slows you down more before you settle into the new rule.
• High error rate on switches: a sign you applied the old rule before catching yourself.

Because the score is a difference between two reaction times, it is naturally noisy. A single run can swing based on a few slow trials, so read your switch cost as a range, not a precise figure. Compare it against your own earlier runs on the same device.

You can shrink your switch cost with practice, but the honest picture is nuanced. Part of the gain comes from getting faster at both tasks overall, and part from reading the cue sooner. Much of it is a practice effect specific to this color-versus-shape setup rather than a broad boost to flexibility in daily life.

• Read the cue first: lock in which question you are answering before you look at the item.
• Do not pre-commit: resist deciding your answer until you know which rule applies.
• Stay even: a steady pace produces a cleaner, lower switch cost than bursts of speed and errors.

Expect limits. There is a ceiling: some switch cost almost always remains, because changing rules is inherently costly. And because the score is a noisy difference, regression to the mean is strong here. An unusually low switch cost on one run is often followed by a higher one, so judge yourself on your average across several attempts.

The classic mistake is answering before reading the cue. If you decide based on the last rule and the rule just switched, you either make an error or catch yourself and lose time. Both inflate your switch cost.

• Deflates your score: racing so fast that switch-trial errors pile up.
• Deflates your score: ignoring the cue and answering on autopilot.
• Makes results misleading: too few trials, which leaves the switch cost dominated by random noise.
• Adds friction: a slow input device that hides your true reaction times.

The fair way to use this test is to run enough trials, keep your device and focus steady, and compare your average switch cost over time. A shrinking gap then reflects real improvement, not a lucky, low-noise run.