ANT: Attention Networks
3 attention networks in one test
About this trainer
ANT is a cued arrow task. On each trial you wait for a fixation cross, sometimes see a cue (a flash above or below where the target will appear), then a central arrow flanked by other arrows, and you press left or right depending only on which way the centre arrow points while ignoring its neighbours. By comparing your reaction times across cue types and flanker types, the test estimates three things separately: alerting, orienting, and resolving conflict.
What it develops
It exercises selective and executive attention, in particular the ability to lock onto a target and suppress distracting information around it, plus the readiness and spatial-orienting side of attention.
History
It grew out of decades of attention research and was published as a single combined task in 2002, fusing two classic paradigms: Posner's spatial cueing and the Eriksen flanker task. Because it is short, language-free and works with children, patients and even monkeys, it spread quickly through cognitive neuroscience, fMRI and genetics labs and now exists in many revised forms.
Who created it — and when
The Attention Network Test was created by Jin Fan, Bruce McCandliss, Tobias Sommer, Amir Raz and Michael Posner, published in 2002 in the Journal of Cognitive Neuroscience. It operationalised the three-network model of attention that Posner and Petersen had proposed in 1990.
How to train
Keep your eyes on the central fixation point and respond only to the middle arrow, never to the flankers. Aim for accuracy first and let speed come with it, because rushing inflates errors on the conflict trials. Don't pre-guess the side; let the arrow tell you, and treat the cue only as a 'something is about to happen' signal.
How long to practise
A full run is roughly 20 to 30 minutes, which is more than a casual daily drill. Short sessions of 10 to 15 minutes a few times a week are enough to stay sharp on the task; there is no evidence that grinding it for hours adds anything.
Evidence base
Evidence is strongest for exactly what you would expect: with practice you get faster and more consistent on this specific cued-flanker task, and the conflict (executive) score is reasonably stable from session to session. The catch is psychometric: the alerting and orienting difference scores are noisy and replicate poorly between sittings, so single readings of them mean little. Claims that training ANT widens general attention, boosts intelligence or holds off cognitive decline are not established, so treat the grand promises with caution.
Recommendations
Use it as a focused warm-up: one clean run with your gaze fixed and your eyes on accuracy, rather than many sloppy speed runs.
FAQ
What do the three scores actually mean?
Each is a difference in reaction time: alerting is how much a warning cue speeds you up, orienting is the benefit of a cue that shows where the target will be, and executive control is how much the conflicting flankers slow you down.
Is the ANT a reliable measure of my attention?
Partly. The executive-control (conflict) score is fairly stable, but the alerting and orienting scores are statistically noisy, so don't over-read a single result, especially for those two.
Will training on the ANT make me more attentive in daily life?
You will clearly improve at the ANT itself, but broad carry-over to everyday focus is not proven. Treat it as targeted practice, not a guaranteed upgrade to your general attention.
Variants
Common variants include the child (fish) version, the ANT-R (revised) with extra cue and flanker timings to probe network interactions, the Attention Network Test for Interactions and Vigilance (ANTI-V) that adds a sustained-attention component, and lateralised versions that present targets to one visual field.