Trail making test with the B minus A difference

The design is nearly a hundred years old and has barely changed, because there is very little in it to change. Two sheets, the same number of circles, the same scattering, the same instruction to work fast without lifting the pen — and one difference, which is that the second sheet asks you to hold two ordered sequences at once and swap between them at every step. The version in clinical use came out of a US Army test battery in the 1940s and reached neuropsychology through Ralph Reitan, who paired it with the rest of what became the Halstead-Reitan battery. Its durability comes from the subtraction rather than from either sheet: part A is a measure of visual search and hand speed, part B is the same measure plus alternation, and B minus A is one of the very few executive quantities you can obtain with a pencil and a stopwatch.

What does not survive the move to a browser is the absolute time, and the reason is a procedural detail rather than a technical one. In the standard administration the examiner does not merely record errors — they interrupt them, name the correct target and return the participant to it, and the stopwatch runs throughout. That correction is part of what the published figures were measuring, so a page where a wrong click is silently refused and the visitor works it out alone is timing something else, before you even consider what a trackpad does to a movement that was designed for a pen. Tombaugh (2004), Trail Making Test A and B: Normative data stratified by age and education, Archives of Clinical Neuropsychology is where the age- and education-stratified tables live for anyone who wants to see what a paper administration produces; not one cell of them is reproduced here, and no figure on this page is scored against them. The difference and the ratio are yours to compare with your own, which is a smaller claim and a true one. The same trade appears on the digit symbol substitution test, where the published version is timed with a pencil for the same reason and the browser version reports its own throughput instead.

Four things go wrong in most web versions of this task. A layout that is the same on every load turns the second attempt into recall. A running timer on screen changes how people work, usually by making them faster and less accurate in a way that lands unevenly across the two boards. Letting a wrong click advance the trail removes the error cost that the paper form charges in seconds, and adding a fixed time penalty per error instead invents a quantity the procedure never had. Worst of the four is the percentile: a table of times by age, applied to a run taken with a mouse on an unknown screen with no examiner present, produces a number that looks authoritative and is simply false. If you want the switching cost measured where the rule itself is hidden rather than printed on the circles, the Wisconsin card sorting test is the paired page; for attention moving across a display rather than between sequences, it is the posner cueing task. Two others charge the same kind of cost in a different currency: the tower of london test counts the moves you had to plan before making the first one, and the balloon analogue risk task keeps a sequence going but never tells you where it ends.

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.

On this page that floor is small enough to ignore and is stated anyway: a board takes tens of seconds and a frame is about a sixtieth of one, so it is roughly a twenty-thousandth of a part A total. It matters in exactly one place — the individual steps drawn under the result, where a fast pair of adjacent circles can be a few hundred milliseconds apart and the frame is then a visible share of the bar.

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 a brain injury, dementia, ADHD or any other condition. Only a qualified professional, working with more than a browser, can make that judgment.