Roxane Cohen Silver began trying to understand how graphic images pollute the human psyche in the terror-struck days after the 9/11 attacks, when sights and sounds of the twin towers crumbling flooded newspapers, television and radio.

The latest work by the professor of psychological science at the University of California at Irvine arrives in the midst of a reckoning with another terrorist attack. The massacre on Friday of 50 people at two mosques in Christchurch, New Zealand, occurred halfway around the globe from Ground Zero, but on a whole other planet of opportunities to view extremist violence.

In 2001, the images of billowing smoke got their oxygen almost exclusively from legacy media. In 2019, the rat-a-tat of gunfire issuing from semiautomatic weapons — and footage of the bloodshed inflicted by the weaponry — became available live on Facebook and other social media platforms.

The grisly events in New Zealand, which reignited debate about the ethics of viewing terrorist propaganda and the responsibilities of technology companies to curtail its circulation, occurred soon after Silver’s newest paper appeared in the American Psychologist, the flagship journal of the American Psychological Association. Her results, published late last month, hold clues about what sort of people are drawn to these images of destruction, as well as the psychological toll the gruesome material may take.

The study, “Who Watches an ISIS Beheading — and Why,” examines engagement with videos depicting the decapitation of journalists James Foley and Steven Sotloff. The videos, each several minutes in length, appeared on the Internet about three weeks apart in 2014.

They shocked the world, enlisting modern methods to achieve barbaric ends. But the paper is the first to quantify how many people actually watched this graphic content, who these people were and what their motivations were for doing so.

In a survey of more than 3,000 U.S. residents that began in 2013 and lasted for three years, Silver and her colleagues at UCI found that 20 percent had watched at least part of a beheading video, while 5 percent said they had watched an entire one.

Survey participants were more likely to say they had watched the beheading videos if they also said they were male, Christian or unemployed. Those who reported watching television more often were also more likely to view the videos. So, too, was fear of future terrorism associated with greater likelihood of watching. Prior mental health conditions were not significant factors, but a lifetime exposure to violence was.