Dot-probe task

Show a pair of images side by side, replace one with a probe, and measure how fast the probe’s location is reported, in one row per trial.

When to use it

Use this for spatial attention paradigms built on the dot-probe logic: two stimuli compete for attention (a threatening and a neutral face, an emotional and a neutral word), then a probe appears at one of the two locations, and the response time to the probe indexes where attention was. Faster responses when the probe replaces the emotional stimulus (congruent trials) indicate attentional bias towards it. The whole structure lives in one trial row: fixation, the pair on one screen, the probe replacing it, and a speeded location response.

Trial file

type	content	stimFormat	stim1	stim2	stim3	stimScreens	presTime	ISI	ITI	stimPos	fixation	keyboard	key	side	congruency	trialOrder	button1
instructions	Press F when the dot appears on the LEFT and J when it appears on the RIGHT. Keep your eyes on the cross.																NEXT
test		.jpg	threat1	neutral1	dot	2+1	500	0	800	-200 0;200 0;-200 0	ITI	f j	f	left	congruent	1: shuffle; maxRun 3 side
test		.jpg	threat1	neutral1	dot	2+1	500	0	800	-200 0;200 0;200 0	ITI	f j	j	left	incongruent	1
test		.jpg	threat2	neutral2	dot	2+1	500	0	800	200 0;-200 0;200 0	ITI	f j	j	right	congruent	1
test		.jpg	threat2	neutral2	dot	2+1	500	0	800	200 0;-200 0;-200 0	ITI	f j	f	right	incongruent	1
test		.jpg	threat3	neutral3	dot	2+1	500	0	800	-200 0;200 0;-200 0	ITI	f j	f	left	congruent	1
test		.jpg	threat3	neutral3	dot	2+1	500	0	800	-200 0;200 0;200 0	ITI	f j	j	left	incongruent	1
test		.jpg	threat4	neutral4	dot	2+1	500	0	800	200 0;-200 0;200 0	ITI	f j	j	right	congruent	1
test		.jpg	threat4	neutral4	dot	2+1	500	0	800	200 0;-200 0;-200 0	ITI	f j	f	right	incongruent	1

How it works

The screen structure is stimScreens = 2+1: three stimuli grouped into two screens, the pair together first, then the probe alone. With the grouping stated there, the timing columns are plain value lists. ISI = 0 is the single gap between the two screens (the probe replaces the pair instantly), and presTime = 500 times the first screen; with one entry fewer than the number of screens, the final screen deliberately has no duration, so the pair shows for 500 ms and the probe stays up until the F or J response.

stimPos is the manipulation itself. One h v segment per stimulus places the threat image at -200 0 (left), the neutral at 200 0 (right), and the probe at whichever of those two offsets the trial calls for: a probe offset matching the threat’s position makes the trial congruent. The side and congruency columns are custom annotation columns; they change nothing on screen and pass through to the results as your analysis labels.

fixation = ITI shows the central cross during the inter-trial interval, disappearing when the pair comes up, so every trial starts from fixation with no extra machinery. The cross needs an interval to show in, which is what ITI = 800 provides; without an ITI value there is no inter-trial period and the cross never appears. key scores the probe’s true location against the F/J response, and the ordering spec 1: shuffle; maxRun 3 side (written on any one row of the group) randomises the trials and stops the threat image from appearing on the same side more than three trials in a row (the side column annotates the threat position; constrain a probeSide column instead if the probe side is what must not streak).

Variations

• Stimulus-onset asynchrony between pair offset and probe onset: change the gap, ISI = 100 inserts a 100 ms blank before the probe.
• Longer or shorter exposure: presTime of 100 for subliminal-style presentation or 1250 for the classic long exposure; the probe screen stays unlimited either way.
• A response deadline: responseWindow = 2000 records timeout and moves on when no response comes within 2 s.
• Posner spatial cueing is the same skeleton with a single cue instead of a pair: stim1 the cue, stim2 the target, stimScreens = 1+1 with ISI = 0 (or a gap for the cue-target interval), and stimPos deciding valid and invalid locations.
• Balance the design in the shuffle: extend the spec to 1: shuffle; balance 8 congruency; maxRun 3 side so any window of 8 consecutive trials stays balanced on congruency as well.