Task Switching and Interruption Recovery Assessment for Interrupt-Heavy WorkThe cost is not the interruption. It is the first minute after it.
Nine sorting streams change the rule in front of you and measure what the first decision after the change is worth.
Everybody knows interruptions are expensive. Almost nothing measures your own price for one.
The rule is printed on the screen the whole way through and it changes three times in every stream. Nothing has to be remembered, and that is deliberate: this is not a memory test. What is measured is whether the instruction you have just been given reaches the very next decision, or whether the one before it gets there first.
The headline is a difference, not a level. Your accuracy on the settled positions is scored, your accuracy on the first position after a change is scored, and the figure is the gap between the two. Somebody accurate everywhere and somebody inaccurate everywhere both land near zero. That is a property a raw accuracy percentage cannot have, and it is why most attention tests tell you about your ability rather than about your cost.
All three sorting rules use the same three bins. That construction is what makes an old-rule answer identifiable afterwards as an old-rule answer, so the report can tell you whether your errors after a change are the previous rule still running or simply errors. Those two have different fixes.
Four more short parts follow and none of them adds to the headline. A coding row, a figure series, a grid span and a fast count are there so the report can say whether a large cost is explained by something more ordinary, and each is reported on its own with a placement rather than a number.
Nothing here is clinical. This is not a diagnostic instrument, it says nothing about any medical condition, and the report says so in its own words rather than in a footnote.
What you walk away with
One number, both spans drawn and written out, and the rung claimed only when the whole band sits inside it.
Accuracy by how many decisions have passed since the change, with every value printed beside its point.
Of the errors you made after a change, how many were the answer the previous rule would have given.
Coding speed, rule-finding, short-sequence span and quantity sense, each measured and each kept out of the figure.
Including what a small switch cost costs you, which is dropping the previous job cleanly when it mattered.
Naming the exact moment your own sitting was weakest, and one small behaviour for it.
Inside your report
Illustrative sample — your report is generated from your own responses.
A rung is claimed only when the whole error band sits inside it. Otherwise the report names the two rungs the result lies across, which is the sentence most reports leave out.
Every value is printed next to its point and repeated as a table underneath, so the curve survives a black and white print.
Built for
- Anybody in interrupt-heavy work: support, operations, clinical, engineering, customer-facing roles
- People deciding how to arrange their own week rather than trying harder at it
- Teams designing handovers, on-call rotations and shared inboxes
- Students and graduates curious about a measurement of their own, at a price that is not a barrier
Find out what an interruption actually costs you
35 exercises across five formats · about 30 minutes · a switch cost with its band, and four things it is not.
Take free · full report ₹249 (incl. GST)
Frequently asked questions
No. It is not a clinical or diagnostic instrument of any kind, it does not screen for anything, and it says nothing about any medical condition. It measures one narrow thing: how much accuracy is lost in the first decisions after a stated rule changes.
No. The rule is printed on the screen the whole way through and you are told when it changes. That is the design. A version that hid the rule would measure memory, which is a different thing and already measured elsewhere.
Anything you did not reach leaves the denominator rather than counting as wrong. A position you never saw is not a position you got wrong, and the report says how many that was.
Because they measure other abilities. Coding speed, rule-finding, span and quantity sense are there so the report can say whether a large switch cost is explained by something more ordinary. Folding them in would produce a number that means none of the five things.
About thirty minutes for thirty-five exercises. ₹249 in India, inclusive of GST, or US$2.99 elsewhere, one time, for the sitting and the full report.
Each one takes a single capability, puts you inside the situations where it is actually tested, and scores your choices against published evidence — with a report designed for that capability alone, not a template. They span hiring, compliance, education, operations and personal skill.
Browse the catalogue →Methodology: Thirty-five exercises across five item formats, none of them a questionnaire and none of them a situational judgement item: nine sorting streams under a rule that flips, seven symbol-coding rows, seven figure series, six grid spans and six quantity estimates. The declared response instruction is performance throughout - every exercise has a right answer and none of them asks what the respondent believes about themselves. Construct statement: it measures how much accuracy is lost in the first decisions after a stated sorting rule changes, and how many decisions it takes to come back. It does not measure intelligence, it is not a clinical or diagnostic instrument of any kind, it says nothing about any medical condition, it does not predict productivity, and it is not a fitness-for-duty check. The four supporting parts measure four other abilities and are reported apart from the headline in words as well as in arithmetic. Scoring is a matched-trial within-person contrast: the positions immediately after a rule change and the settled positions later in the same run are scored separately, each chance-corrected against that position's own authored answer priors rather than against a coin, and the headline is the difference between the two. Somebody uniformly accurate and somebody uniformly inaccurate both land near zero cost, which is the property a raw accuracy percentage does not have. About three positions in ten were built so the old rule and the new rule point at the same bin, so that never repeating the previous answer is worth nothing. Because all three rules key onto the same three bins, the answer the previous rule would have given is identifiable, and the share of errors that are old-rule answers is reported beside the cost. Constructs and sources: task switching and the reconfiguration account (Monsell 2003; Rogers and Monsell 1995; Meiran 1996); task-set inertia and proactive interference from the abandoned rule (Allport, Styles and Hsieh 1994; Wylie and Allport 2000); the review literature on switch cost, cue-target interval and congruency (Kiesel and colleagues 2010; Vandierendonck, Liefooghe and Verbruggen 2010); perseverative responding after a rule change in card-sorting paradigms (Berg 1948; Milner 1963); shifting as a separable executive function (Miyake and colleagues 2000); resumption lag after workplace interruption (Mark and colleagues on interrupted work; Altmann and Trafton's memory for goals); digit-symbol substitution as a measure of clerical speed (Wechsler; Joy, Kaplan and Fein on what the task actually measures); block-tapping span (Corsi 1972; Kessels and colleagues 2000); figural series and rule induction (Carpenter, Just and Shell 1990); approximate number sense and its measurement (Halberda and Feigenson); and item-writing guidance from Haladyna, Downing and Rodriguez. All stimuli and keys are original works generated for this instrument. No commercial or clinical test's items, name, norms or scoring is reproduced or implied, and no affiliation with any test publisher exists or is suggested.