Trail making test: method

What it measures

Part A: visual search and psychomotor speed. Part B adds alternating between two sequences (set shifting). B minus A removes most of the search and motor part and leaves the switching cost (Sánchez-Cubillo et al. 2009).

Procedure

Circles are scattered at random on a 4 x 3 board (radius 0.2 board units, centres at least 0.62 units apart); every run has new layouts. On narrow screens the board is turned on its side. Before each part an untimed 8-circle sample (default for new players, skippable). Part A: 20 circles numbered 1 to 20. Part B: 20 circles, 1 to 10 and A to J, to be tapped alternately. The player taps or clicks the circles in order. A wrong tap flashes red and is counted, and the trail continues from the last right circle; taps on circles already done are ignored. Each part is timed from the moment its circles appear to the tap on the last circle. A part longer than 5 minutes ends the run. Leaving the browser tab ends the run.

Parameters

Scoring

Time A and time B = from the circles appearing to the last circle (shown in seconds). Primary: B minus A, in seconds; lower is better. Also: B/A ratio and errors per part (taps on any later circle than the next correct one). No player comparison is shown.

Valid-run rules

Both parts are needed and each must reach its last circle. The server rebuilds the layouts from the run's seed; tap times must not go backwards or pass 5 minutes. The server rebuilds every run from the raw log (and the run's seed where there is one) and computes the score itself; anything the page says about right or wrong is ignored. A log that breaks the game's rules is rejected, and nothing is stored. Stored runs can carry quality flags that are never shown and never block a result: 10 or more runs from one IP address, a missing visitor id, tab switches, and automation signals (automated browser, impossible speed, no input events, machine-regular timing).

Differences from the classic task

The paper test has 25 circles per part (part B: 1 to 13 and A to L) on a fixed printed sheet, drawn with a pencil; the examiner points out errors, which count only through the time lost. Here there are 20 circles, random layouts each run, tapping instead of drawing, errors counted separately, and the board size depends on the screen. Published paper norms do not apply.

Limitations

Finger versus mouse, screen size and layout luck change the times. Part B uses Latin letters in every language, which is harder for people who rarely use the Latin alphabet. Practice shortens times. Unsupervised setting. Not a clinical test.

Languages and review status

Numbers 1–20 and Latin letters A–J on the board in every language (the board is not localised); instructions in all site languages (English plus 99 other languages).