As many office workers transitioned to working from home last year, there has been an unprecedented rise in the number of pet adoptions. And as any dog or cat owner will attest, caring for a pet comes with many challenges. Your cat might block your view of your computer screen all day, but when it’s time to take him to the vet, he’s nowhere to be found. And of course, after scanning your house, you finally find Felix hiding in the pile of laundry you walked past several times already.
While this may seem like a unique challenge, this process of visual search—looking for items in clutter—is ubiquitous in everyday tasks, such as looking for a specific book on your shelf, finding the exit sign as you drive down the highway, or grabbing a pen off your desk. In a recent paper in Psychonomic Bulletin & Review, Jeremy Wolfe has refined a model that describes the perceptual and cognitive process of how we find what we’re looking for (and in doing so, maybe also make it easier to find his newly adopted kittens).
Visual search is necessary because the world carries a lot of visual information, and there is a limited amount that we can attend to at a given moment in time. Looking for the cat in the photo has made it pretty clear that it takes some time to figure out what’s going on in the scene. There is some information that you can get preattentively – basically, in a single glance. For example, you can tell that this is an indoor scene, and that there’s a set of bookshelves. Finding the blue vase, however, requires going through and selectively attending to the individual items on the shelves.
In visual search, how do you determine what to select next? Wolfe’s model, called Guided Search, is based on the idea of guidance: you can restrict your attention to a subset of the items based on information about the item’s basic features (such as color, size, etc.). If you know that the vase you’re looking for is small and blue, you don’t need to waste your time scanning the red or green items, or the larger objects (i.e., chairs, boxes) in the scene. In other words, information about these features (the intersection of “blue stuff” and “small stuff”) can be used to guide sequential shifts of attention. This is an example of “top-down” guidance – guidance that is based on an item’s basic features as represented in memory. Guidance can also be based on “bottom-up” information – i.e., information that is salient or stands out from its surroundings.
Over the years, Wolfe has refined this basic model to account for other types of guidance (there have been quite a few updates since the first version in 1989, with Guided Search 2.0 being, perhaps, the best known). The latest version, Guided Search 6.0, adds three new forms of guidance:
- from value (what are these targets worth to you),
- from history (what have you attended to before), and,
- most importantly, from the structure and meaning of the scene as a whole.
Through your experience, you know that when looking for your car keys in the house, they’re unlikely to be on the stove, or suspended from the ceiling. You can restrict your attention to those locations where it’s possible or sensible to find the item. Or, to take this example from the paper, where might you expect to find a sheep?
In Guided Search 6.0, Wolfe also discusses what your visual system might do once you’ve attended to an object. Once you’ve attended to it, what happens to that representation? To start with, that representation is held in working memory, which has limited capacity and is compared to the search template (what you’re looking for), which is held in Activated Long Term Memory. If they match, great, you’ve found the blue vase (or the cat). If not, you select another item from the scene, attend to it, load it into working memory and try again. You keep doing this until either you find what you’re looking for, or give up (maybe because the cat moved while you weren’t looking).
Importantly, you can’t look everywhere at once in the scene, and Guided Search 6.0 incorporates Functional Visual Fields, which define the spatial limits of where you can look. In other words, they describe what you can identify without moving your eyes and where you can look next. Using an example from the paper, keep your eyes on the star in the image below, while searching for the oval shapes. You may only be able to identify the blue oval at 2 o’clock and the purple one at 4 o’clock (and miss the other two ovals). Search difficulty can vary depending on how near or far your targets are, and how different they are from their surroundings. It’s not that you only see a small region of the scene at once, it’s that it may be harder for you to find things you’re looking for if they’re far away from where you’re looking. The introduction of Functional Visual Fields takes into account these types of spatial limits in search.
In sum, Guided Search 6.0 provides a framework for understanding how visual search operates, and by adding several new forms of guidance and the incorporation of spatial factors, can help explain why you can (or can’t) find what you’re looking for. And while Guided Search 6.0 can account for many laboratory results on search, Wolfe points out that search outside the laboratory can be much more complex. Searching for the ovals in the image above is quite different from ‘extended’ searches, where a radiologist may be looking for cancer in a CT scan, or a search where navigation is involved (like looking for your cat around the house). Further work will be needed to determine how well these laboratory studies apply to these difficult (and sometimes hairy!) real-world tasks.
Featured Psychonomic Society article:
Wolfe, J. M. (2021). Guided Search 6.0: An updated model of visual search. Psychonomic Bulletin & Review, DOI: https://doi.org/10.3758/s13423-020-01859-9.