Trail Making Test

Part A is a search and pointing task. Find the next number, move, click. Part B adds one thing: you swap between numbers and letters. That single change usually doubles your time.

Both parts involve the same searching and the same pointing. Subtract one from the other and all that shared work cancels out, leaving the cost of switching.

Sánchez-Cubillo and colleagues checked what part B really leans on. They found working memory and self-control alongside hand-eye speed. That is why B on its own measures nothing in particular, and B minus A measures something.

The trail making test began life in a US Army test set. Ralph Reitan made it famous. In 1958 he reported that it told people with brain damage apart from people without.

It has been in clinical use ever since, mostly inside the Halstead-Reitan set. It is still one of the most given neuropsychological tests in the world.

That history is why its reference figures are unusually good. Tombaugh's 2004 study broke results down by age and by years of education across a large sample. Both matter enormously. Education moves part B times more than a decade of ageing does. Any average quoted for this test without an age and an education level is close to meaningless.

The paper version is drawn with a pencil that never leaves the page. The line is unbroken, and a tester fixes your mistakes by guiding you back. On a screen you click separate targets, so there is no drawn line, no hand travelling across paper and no tester.

This version handles mistakes the way a paper session does, which is the part most online versions get wrong. A wrong click is refused and flagged, the clock keeps running, and you carry on from the last correct circle. The time you spend recovering is therefore inside your score, exactly as it is on paper.

Circle positions are shuffled every run, with a minimum gap between them, so nobody memorises a layout.

A part B time near 60 seconds sits in the middle of the reference distribution for younger adults. The range is wide and it depends heavily on age.

So treat any comparison against a published clinical table with suspicion unless it matches your age band and your education level. Remember too that a mouse is a different instrument from a pencil.

The number to watch on yourself is B minus A. If both times rise together, you probably just got slower at pointing. If B rises while A holds steady, your switching cost went up, and that is the more interesting event.

For a second angle on the same ability, the Stroop test measures interference rather than switching. What the Stroop effect actually shows explains why subtracting one score from another beats a single total. For how steeply speed measures move with age, see average reaction time by age.