Visual Search Task: Set Size, Theory and Training Studies

What is a visual search task? How a visual search experiment works

In a classic lab version, described by Jeremy Wolfe in 2021, you report whether a target is present among other items, and it is there on about half of the trials. Researchers change how many items appear, called the set size, and record reaction time and errors for each.

The set-size effect is the main finding. For some searches, adding distractors makes little difference: the target seems to pop out, and the reaction time slope is near zero milliseconds per item. For others, search time rises roughly in a straight line as items are added. Wolfe gives the example of a digital 2 among digital 5s, where slopes are around 90 milliseconds per item when the target is absent and about half that when it is present, because on average you check about half the items before finding it.

Feature vs conjunction search and guided search

Ulric Neisser published visual scanning experiments in 1963, and Wolfe's 2021 review says the main search patterns were uncovered in the 1960s and 1970s. Wolfe writes that those patterns formed the basis of the feature integration theory, set out by Anne Treisman and Garry Gelade in 1980.

As later researchers describe it, the theory held that basic features such as color and orientation are processed in parallel across the whole view, so a target defined by one feature pops out. A target defined by a conjunction, such as a red vertical bar among red horizontal and green vertical bars, needs attention to bind its features, so items are checked one by one.

Guided search was a response. In 1989, Wolfe, Cave and Franzel tested conjunction searches and found that for many unpracticed people, slopes were too shallow to fit strictly serial search. Searches for triple conjunctions were easier still. They proposed that parallel feature processing guides attention toward likely targets. The 2021 version, Guided Search 6.0, names five sources of guidance: what you are looking for, what stands out, recent history, reward, and the structure and meaning of a scene.

Visual search in daily life and online visual search tests

Wolfe's examples of everyday search include finding the fork at dinner, spotting an exit sign while driving, radiologists reading scans and security officers checking bags.

Screening jobs like these are searches for rare targets, which is where the prevalence effect comes in. In a 2005 Nature paper, Wolfe and colleagues reported that when targets are rare, people often miss them when they do appear. A 2014 study by Mitroff and Biggs used millions of trials from the Airport Scanner smartphone game and found that detection of extremely rare items was very low.

Clinicians use search-based paper tests too. Cancellation tests ask you to cross out every target on a page. In a 2001 study of 35 people with spatial neglect, the letter cancellation and bells tests each missed 6% of cases, while line bisection missed 40%.

An online visual search test can show the set-size effect, but scores from different sites are not comparable. A 2021 study of online experiment platforms found reasonable timing overall, but found that response times were recorded with an average lag of about 80 milliseconds, reaching 100 milliseconds on some set-ups. We did not find validated norms for online visual search tests.

Visual search training: what practice studies found

People get faster at search tasks they practice. In 1995, Sireteanu and Rettenbach reported that some searches that start out item-by-item became pop-out searches after a few hundred trials. Their 2000 follow-up found the learning was retained over several months and transferred to other search tasks, other locations in the visual field and between the two eyes. Conjunction search also became more efficient with long practice.

In a 1997 study, Ahissar and Hochstein found that learning on easy versions of a pop-out task carried over to new orientations and positions, while learning on harder versions became more specific to the trained orientation and position.

A 2013 study used a visual search game as the active placebo for working memory training; our working memory games guide covers it. A 2016 review led by Daniel Simons found extensive evidence that people improve on trained tasks and little for daily life; our visual brain training guide summarizes it.

Visual scanning exercises after stroke: hemianopia and neglect

After a stroke, some people lose the same half of the visual field in both eyes (homonymous hemianopia), and some stop attending to one side of space (spatial neglect). Visual scanning training teaches deliberate eye movements toward the affected side, often with search displays.

A 2019 Cochrane review of visual field loss after stroke found four studies (193 participants) of scanning training against a control or placebo. It found low-quality evidence that scanning training improved quality of life on a vision questionnaire, and low or very low quality evidence of no effect on the visual field, extended activities of daily living, reading or scanning ability. A 2021 Cochrane review of neglect included 65 trials with 1,951 participants and rated the evidence for visual interventions on lasting daily function as very uncertain.

Brief-exposure search and the useful field of view

In brief-exposure search, the display flashes and disappears, so you cannot move your eyes around it. That links search to the useful field of view (UFOV), the area from which you can take in information in a single glance. Karlene Ball and colleagues studied it in a 1988 paper titled "Age and visual search: expanding the useful field of view". They found the field was smaller in older adults and partly recovered with practice.

UFOV training was later tested in the ACTIVE trial of older adults. Our guides to visual brain training and peripheral vision training cover ACTIVE and action video game studies of locating brief targets away from center.

Circle Spot, C Count and Middle Match: brief-flash search games

Circle Spot and C Count are modeled on brief-exposure visual search, and Middle Match is similar to the useful field of view task. All three ask you to find a target among distractors in a display that flashes.

• Circle Spot is modeled on brief-exposure search with localization. One O appears among Cs turned in different directions on a grid, then everything disappears, and you tap where the O was. As levels rise, more symbols appear, the Cs get smaller and the flash gets shorter, so set size and exposure both change. Whether the O pops out among the Cs was not tested.
• C Count is modeled on search with counting. A card shows a target direction, then cubes carrying Cs rise from water for a short time. You count the Cs whose gap points the target way and choose the answer from four numbers. More cubes appear as levels rise, and some levels give you less time to count them.
• Middle Match is similar to the useful field of view task, except that the center C sets which ring C you look for. A center C and a ring of Cs flash together; exactly one ring C faces the same way as the center one, and you tap where it was. The ring gets more Cs, a wider radius and a shorter flash as levels rise.
• All three have an optional binocular mode, with red-blue glasses on a phone, tablet or computer, or built in on Meta Quest. They score points and adjust the level after each answer, rather than recording reaction time slopes. Any score is a practice result, not a clinical measurement.

Where Optics Trainer fits

Optics Trainer is a set of vision and brain training games for iPhone, iPad, Android, web browsers and Meta Quest. Circle Spot and C Count are modeled on the visual search tasks described above, Middle Match is similar to the useful field of view task, and all three challenge you to find targets in a brief flash.

These games were not tested in any of the studies described here. They are not tests or measures, give no reaction time slope or norm, and are not a treatment or diagnosis. They are not FDA cleared and are not an alternative to scanning training or rehabilitation prescribed after a stroke. We make no claim that they change attention, peripheral vision, driving, reading or daily life. For clinical measurement of eye movements, see Optics Tracker on our clinic site.

Optics Trainer is not a substitute for professional diagnosis or prescribed care. Consult your eye care professional for guidance about your specific condition. The phone and tablet apps start with a 7-day free trial, and a free web account includes 3 plays a day. After that, mobile and web access is $14.99 a month or $149.99 a year, and VR access is $19.99 a month or $199 a year after a 7-day free trial.

Frequently asked questions

Sources

1. Treisman AM, Gelade G. A feature-integration theory of attention. Cogn Psychol. 1980
2. Neisser U. Decision-time without reaction-time: experiments in visual scanning. Am J Psychol. 1963
3. Wolfe JM, Cave KR, Franzel SL. Guided search: an alternative to the feature integration model for visual search. J Exp Psychol Hum Percept Perform. 1989
4. Wolfe JM. Guided Search 2.0: a revised model of visual search. Psychon Bull Rev. 1994
5. Wolfe JM, Horowitz TS. Five factors that guide attention in visual search. Nat Hum Behav. 2017
6. Wolfe JM. Visual search: how do we find what we are looking for? Annu Rev Vis Sci. 2020
7. Wolfe JM. Guided Search 6.0: an updated model of visual search. Psychon Bull Rev. 2021
8. Wolfe JM, Horowitz TS, Kenner NM. Rare items often missed in visual searches. Nature. 2005
9. Mitroff SR, Biggs AT. The ultra-rare-item effect: visual search for exceedingly rare items is highly susceptible to error. Psychol Sci. 2014
10. Ferber S, Karnath HO. How to assess spatial neglect: line bisection or cancellation tasks? J Clin Exp Neuropsychol. 2001
11. Anwyl-Irvine A, et al. Realistic precision and accuracy of online experiment platforms, web browsers, and devices. Behav Res Methods. 2021
12. Sireteanu R, Rettenbach R. Perceptual learning in visual search: fast, enduring, but non-specific. Vision Res. 1995
13. Sireteanu R, Rettenbach R. Perceptual learning in visual search generalizes over tasks, locations, and eyes. Vision Res. 2000
14. Ahissar M, Hochstein S. Task difficulty and the specificity of perceptual learning. Nature. 1997
15. Pollock A, et al. Interventions for visual field defects in people with stroke. Cochrane Database Syst Rev. 2019
16. Longley V, et al. Non-pharmacological interventions for spatial neglect or inattention following stroke and other non-progressive brain injury. Cochrane Database Syst Rev. 2021
17. Rowe FJ, et al. A randomized controlled trial of Scanning Eye trAining as a Rehabilitation Choice for Hemianopia after stroke (SEARCH). Int J Stroke. 2025
18. de Haan GA, et al. The effects of compensatory scanning training on mobility in patients with homonymous visual field defects: a randomized controlled trial. PLoS One. 2015
19. Roth T, et al. Comparing explorative saccade and flicker training in hemianopia: a randomized controlled study. Neurology. 2009
20. Pambakian AL, et al. Saccadic visual search training: a treatment for patients with homonymous hemianopia. J Neurol Neurosurg Psychiatry. 2004
21. Ball KK, et al. Age and visual search: expanding the useful field of view. J Opt Soc Am A. 1988

Related guides

• Brain Games Backed by Science? The Lab Tasks Behind Them
• Schulte Table: Test, Average Time and What Studies Show
• Peripheral Vision Training: Games, Apps and the Research
• Multiple Object Tracking: The Task, Tests and Training