Cued rule task
Follow an explicit cue, apply a fixed left/right mapping, and compare repeat with switch trials without collapsing them into an opaque score.
Read the cue, then switch between fixed parity, magnitude, color, or shape rules.
Read the rule cue first, then use the fixed left/right mapping. F = left; J = right.
Current rule
Odd or even
first
RT switch cost = median correct RT on switch trials − median correct RT on repeat trials; accuracy cost = repeat accuracy − switch accuracy.
Overall accuracy
0%
Repeat accuracy
0%
Switch accuracy
0%
Overall median correct RT
—
Repeat median correct RT
—
Switch median correct RT
—
RT switch cost
Not enough comparable trials
Accuracy switch cost
Not enough comparable trials
Errors
0
Timeouts
0
Anticipatory responses
0
No validity flags were detected
Each trial presents a cue and a stimulus with several possible features. The cue tells you whether to judge odd versus even, below versus above five, or—in Advanced and Custom modes—color or shape. The left and right mappings stay fixed throughout the block.
Completed summaries are stored locally in this browser. Accuracy, correct-response medians, timeouts, and switch costs describe this exact rule set and device. The page is not a standardized assessment of cognitive flexibility or a clinical test.
A switch occurs when the current cue names a different judgment rule from the preceding trial. The same number can therefore require different answers: 3 is odd under the parity rule and below 5 under the magnitude rule. Advanced and Custom modes can add either a color or shape rule that ignores the number.
The cue remains visible with the stimulus. This is a cued task-switching design, not a hidden-rule guessing game. Button mappings are shown before and during the block and do not change trial by trial.
A repeat trial uses the same rule as the immediately previous trial. A switch trial uses a different rule. The first trial has no predecessor and is marked first rather than being forced into either comparison group.
Seeded generation balances the sequence around the chosen switch probability and prevents very long runs of one rule. Formal sessions use a fresh random seed, while automated tests use a fixed seed to reproduce classification and scoring.
RT switch cost equals median correct response time on switch trials minus median correct response time on repeat trials. Accuracy switch cost equals repeat accuracy minus switch accuracy. Errors and timeouts remain visible and are not deleted before calculation.
The page shows a cost only when both groups contain at least three trials. Otherwise it says that there are not enough comparable trials. Correct-only RT prevents a fast wrong response from being presented as efficient performance, while accuracy remains alongside speed.
A positive RT cost means correct switch responses were slower in this block; a negative value means their median was faster. Either direction can occur with small samples, different rule mixtures, deliberate caution, or random variation. It is not a graded ability level.
Accuracy can trade against speed. Someone may slow down on switch trials to preserve accuracy, or answer quickly with more errors. Read repeat and switch accuracy, medians, timeouts, and trial counts together.
Practice makes the key mapping and cue vocabulary more familiar, so later blocks are not independent of earlier ones. A shorter deadline, a third rule, or a different switch proportion also changes the demands. Compare only sessions with matching settings.
The page flags responses below 150 ms as anticipatory because they are unlikely to reflect a complete cue-based decision. Such responses remain in the error and timing record rather than being silently discarded.
Stroop and Flanker tasks usually keep the relevant rule stable while introducing conflict from an irrelevant feature or surrounding items. Here, the relevant judgment itself changes between trials and is announced by a cue.
Task switching is one experimental way to study rule reconfiguration, but one short webpage cannot measure cognitive flexibility as a whole. Device latency, preparation time, familiarity, and the particular rules all limit interpretation.
Monsell, S. (2003). “Task switching.” Trends in Cognitive Sciences, 7(3), 134–140. https://doi.org/10.1016/S1364-6613(03)00028-7
These questions stay tied to the visible content on this page.
Continue with tools that pair naturally with this page.
Answer color-word, reverse, or shape rules and review accuracy, median timing by condition, and the page-defined interference difference.
Start
Complete congruent, incongruent, and neutral Flanker trials, then compare condition accuracy, median response time, and interference.
Start
Run a 1-, 2-, or 3-back position-and-letter task and review hits, misses, false alarms, and task accuracy.
Start