Roughly 16 years ago, two researchers wrote about a paradox they observed at the Metropolitan Museum of Art.

Lisa F. Smith and Jeffrey K. Smith—now both professors at the University of Otago—had been studying how visitors experience museums for years. They focused on the way museum-goers thought about their jaunts through white-walled institutions and extolled the cognitive pleasure that came with looking at art. The Smiths found that visitors professed to love museums, describing the experience as “incredible,” “breathtaking,” “outstanding,” and “a thrill of a lifetime.”

But they also noticed something else: People don’t spend that much time looking at art. So, they wondered, in what would become a seminal study, published in 2001, “How can people be so deeply moved by works of art that they have viewed so briefly?”

To answer, they took a step back and asked another question: How much time were the people strolling the halls of the Met spending looking at art? Working with a volunteer, the couple monitored 150 people as they looked at six paintings from the museum’s collection, including famous works like Washington Crossing the Delaware (1851) by and the The Card Players (1890–92) by .

They found that the mean time spent looking a painting was 27.2 seconds, while the median time was 17 seconds, and the longest time was 3 minutes, 48 seconds—recorded as one person observed ’s Aristotle with a Bust of Homer (1653). So, long before the age of the iPhone (and digitally stunted attention spans), Smith and Smith concluded that “a [museum] visit is not characterized by long looks at a few works of art; it is characterized by brief looks at many works of art.”

Today, the question of how long people look at artworks in a museum remains an interesting one—especially to the Smiths. In February 2016, the pair published another, larger study (with more artworks and more people tracked) alongside Pablo P. L. Tinio of Montclair State University, based on research conducted at the the Art Institute of Chicago (AIC). They found “remarkably similar” results as they had with the 2001 study, with the mean time spent looking at a work coming in at 28.63 seconds. The median time had increased to 21 seconds (the mode remained unchanged at 10 seconds).

These results suggest that cell phones haven’t changed the amount of time people spend in front of art that drastically. The big difference the study found was the birth of selfies—or “arties,” selfies taken with artworks, as the authors dubbed them. Though they hadn’t initially planned to measure arties, the authors witnessed the “selfie phenomenon” on the first day of the new study and decided to begin tracking it. Of the 356 observations recorded by the study’s authors after they began tracking these art-selfies, they found that approximately 35 percent involved “arties.” Two people were taking so many “arties” that they had to be excluded from the study, because they weren’t even looking at the art.

Another study published in early 2017, titled “Art Perception in the Museum: How We Spend Time and Space in Art Exhibitions,” sheds even more light on the issue. Researcher Claus-Christian Carbon looked at how much time visitors spend looking at art in a smaller exhibition space that bans photography. After all, visitors to the AIC and the Met are often out-of-towners who are keen to see the collection highlights before catching a flight home or booking it to another tourist destination.

“People are often in a ‘Met mode,’” as Carbon, a professor of psychology at the University of Bamberg, said to me. “They want to see a lot of things.”

His study was focused on six paintings by within a temporary art exhibition devoted to the artist. Observers monitored how long viewers spent with the six works, out of a total of 28 in the show. Unlike the seminal 2001 study by the Smiths, Carbon’s study measured if individuals returned to spend more time with a work after they had already looked at it. He also measured the viewer’s distance from the work.

The median time visitors spent looking at Richter paintings was 25.4 seconds—comparable but certainly longer than both of Smith and Smith’s studies. Interestingly, Carbon’s study found that there was a 51 percent probability that a visitor would return to an artwork at least once, and those who did so had spent 12 seconds less, on average, looking at the piece originally, compared to those who only looked once. This tendency increased the overall time spent looking at a given work of art in the study.

The data Carbon collected of visitors’ viewing distances suggested that people tend to adjust their distance depending on the size of the work (an intuitive finding).

But all of this is significant because a great deal of research on how people perceive art, such has how people’s eyes move across a canvas, occurs through computers in a lab, not actual artworks in a museum. This includes some of Carbon’s own work. The conditions of viewing art on a screen in a lab are generally quite different than at a museum, as his work has demonstrated.

Does this mean that all lab research on art is useless? Absolutely not. Combining lab and field research is useful. “Together, you can really get a comprehensive picture of what we call the art experience,” Carbon said.

As I’ve written about before, there is no “correct” amount of time you need to spend looking at a work of art. It’s completely valid to spend three hours in front of a single work, but it’s also fair to quickly glance at all of the masterpieces in a major museum, or every work in an exhibition. People spend a number of hours watching an opera, the Smiths noted—and that same total time can amount to very little when divided among dozens of artworks in a major retrospective exhibition.

So, if you go to a museum and spend one whole minute looking at an artwork, give yourself a pat on the back. You’re above average.