Numeracy test made of labels, bills and shelf tickets

The international adult-skills surveys treat numeracy as the ability to work with quantitative information as it arrives — embedded in a document, a diagram, a price, a schedule, with irrelevant quantities alongside it — rather than as school mathematics retested on grown-ups. The framework behind that definition is published: OECD (2021), The Assessment Frameworks for Cycle 2 of the Programme for the International Assessment of Adult Competencies, OECD Publishing. Two consequences follow for a page like this one. The first is that items have to be situations rather than expressions, which is why every card here is a label, a bill or a receipt. The second is that the noise has to stay in. An adult numeracy item without surplus figures is an arithmetic item, and the surveys put the surplus in on purpose.

What the format then exposes is a split that a single score hides. There are two ways to get a numeracy item wrong: build the wrong model and compute it correctly, or build the right model and slip on the sum. They look identical in a mark scheme and they are entirely different problems. The first shows up as a factor of a thousand because the column heading said thousands, as a discount taken off the crossed-out price, as a standing charge left out of a bill, as gallons per mile where miles per gallon was wanted. The second shows up as a digit. Reversing the two remedies wastes people’s time: someone whose losses are all modeling errors will not be helped by arithmetic practice, and this is the reason the page asks you to tap the figures you are using before you answer.

The instrument is freely available and designed to sort educated samples into quartiles, but this file could not confirm the band boundaries with enough confidence to print them, and a quartile boundary one item wrong misclassifies a quarter of visitors. The instrument itself is Cokely, Galesic, Schulz, Ghazal & Garcia-Retamero (2012), Measuring risk literacy: The Berlin Numeracy Test, Judgment and Decision Making. So this page reports a count out of ten and never a quartile. If you want the same underlying skill in the format employers use — tables, charts, a calculator, a clock — the numerical reasoning test is that page, and the math aptitude test strips the situation away entirely to ask whether the quantities themselves make sense to you. Where the loss turns out to be the sums rather than the setup, the arithmetic drill measures that directly and the sprint turns it into one number you can chase. Adults arriving here from a job application often need the reading half of the same assessment too, which is what the English level test covers.

A reaction time here is the interval between the frame that painted the stimulus and the timestamp the browser attached to your key, both read from the same monotonic clock. What neither can see is the display pipeline behind it, so on a 60 Hz screen roughly 16 ms of every figure below is the machine rather than you. That is the timing floor: two numbers closer together than that are the same number, and this page reports no precision it cannot support.

None of that floor is visible in the figures on this page, and it is stated because the page does report seconds. A situation here takes twenty seconds or two minutes, so the medians are given to a tenth of a second and the frame time is four orders of magnitude below the thing being compared. What can move a family median is you rereading a card, which is a real event and belongs in the number.

This is a measurement exercise, not a clinical assessment. It reports what you did on this page against a stated reference and nothing more — it cannot establish dyscalculia, or a numeracy level on any national or international scale. Only a qualified professional, working with more than a browser, can make that judgment.