Researchers track eye movements to improve visual searches

"We're looking at how working with a teammate affects behavior," said Michael Hout, assistant professor of psychology and lead investigator on the project. "There are certain aspects of search that might be benefited more from working with a partner, relative to other aspects."

So far, researchers found it to be beneficial for subjects to work together on exhaustive searches. In other words, if there is a display -- like an X-ray -- with multiple items to find, working with a partner is better than working alone. However, in terms of finding things very quickly, working in pairs showed less of an advantage.

Consider a Transportation Security Administration screener looking for belongings that make a bag prohibited: It's best to find one suspicious item on the X-ray as quickly as possible, because the bag will then be removed from the conveyer belt and searched manually for other potentially dangerous items, Hout explained.

"Whereas, with a radiologist, just because you found one tumor doesn't mean that there's not a second one there -- you don't have an inclination to search quickly," he said. "In radiology, you want to search slowly and make sure you don't miss anything."

Hout and his team mimic these high-stakes visual search scenarios in the lab to gauge the performances of both independent searchers and search pairs.

For these studies, Hout said, the subject pool typically consists of volunteers from NMSU's introductory psychology classes, as evidence suggests that the way adult human brains function in terms of information processing changes very little across levels of socioeconomic status or age.

The computer-based experiments allow subjects to tap on a touch-screen monitor to indicate that they've found something or that they've finished searching. To better understand the volunteer's visual search paths, the group uses a desktop eye tracker, which photographs a subject's eyes every millisecond -- a thousand times per second.

"We present different objects on a screen, and we have people search for a target object," said Arryn Robbins, a doctoral student in Hout's lab. "The eye tracker gives us information on where people look."

The tool adds a lot of flexibility and strength to data, Hout explained.

"We can make much more advanced hypotheses about what people are doing when they look for things in their environment," he said.

Like many aspects of cognition, or methods the brain uses to acquire information, visual searches are often conducted so effortlessly that the ability can be taken for granted.

"You don't really think about it when you go looking for your keys in the morning, or your wallet, or whatever the case may be," Hout said. "There's actually a whole lot of information processing going on, outside of conscious awareness; that's the sort of thing that we study here -- the machinery that's going on when you're doing these kinds of tasks."

Currently, Hout and a collaborator from Louisiana State University's Department of Psychology, Megan Papesh, have a grant proposal in review by the TSA detailing the lab's preliminary visual search findings and suggesting new methods for creating an automated training procedure for TSA screeners.

"There's a good bit of research that suggests people who are experts in particular tasks exhibit different eye movements than non-experts," Hout said.

For instance, the scan paths of practiced visual searchers may exhibit calmer and more measured eye movements than newer professionals, whose visual scans appear much more erratic.

By assembling the eye movements of experts and novices across visual search tasks using portable eye trackers, Hout's team plans to create a training tool to help professional visual searchers improve their search abilities.

While the realm of high-stakes visual search has been broadly studied, there hasn't been much research on improvement strategies for low-stakes searches, Hout explained. So, for now, for those who can't find their keys in the morning, consider checking your hands to see if you're already holding them.