Walt Disney PicturesWalt Disney PicturesA study reported this week in JAMA Pediatrics found that children who watched a movie excerpt in which characters used firearms were more interested in playing with a handgun than children who watched an expurgated version from which images of firearms had been excised. The researchers think their findings could be important in reducing fatal firearm accidents involving children. Jenny Anderson, who wrote a Quartzstory about the study, thinks it tells us something about the roots of gun violence. Here is why they are wrong.

In the experiment, Wittenberg University communication researcher Kelly Dillon and Ohio State psychologist Brad Bushman used 20-minute excerpts from two PG-rated movies, National Treasure and The Rocketeer. They randomly assigned 52 pairs of 8-to-12-year-olds to watch either the original scenes or the expurgated, gun-free version, then let the kids play for 20 minutes in a room that contained a disabled handgun in a drawer as well as various toys and games. The subjects were monitored by a video camera and an infrared sensor in the gun that recorded trigger pulls.

On average, the subjects who watched movie excerpts featuring firearms held the handgun longer and pulled the trigger more than the subjects who watched the firearm-free versions. The difference in handling time was not statistically significant after Dillon and Bushman adjusted the data to take into account potential confounding variables such as sex, age, aggressiveness, and attitudes toward guns. But the difference in trigger pulls was robust and dramatic. Whether it has any practical significance is another matter.

Dillon and Bushman worry that children who see guns in movies will be more likely to play with them in real life, with potentially fatal consequences. The implication is that more guns in the movies kids see will mean more fatal gun accidents. But there is no evidence of such a correlation. To the contrary, unintentional firearm fatalities involving children have been falling for decades even though, according to Dillon and Bushman, “gun violence in movies is increasing, especially in movies that target younger viewers.”

Dillon and Bushman cite a study that found the depiction of guns in popular PG-13 movies “has more than doubled since 1985” and a follow-up study that found “the amount of gun violence in PG-13 films continued to increase through 2015.” During that period, according to the U.S. Centers for Disease Control and Prevention, accidental gun deaths involving children 14 or younger fell by 83 percent, from 278 in 1985 to 48 in 2015.

Reasonable people may disagree about the best way to make further progress in reducing the number of children accidentally killed by guns. But even if we assume that the behavior observed in the highly artificial conditions of Dillon and Bushman’s experiment carries over into the real world, securing guns so that children cannot play with them seems like a much more practical approach than restricting their movie viewing or banning guns from PG-13 movies.

While Dillon and Bushman focus on accidental deaths, Jenny Anderson worries in her Quartz piece that seeing guns in movies will make kids “more violent.” Again, there is no evidence that is happening. Even after increases in 2015 and 2016, the violent crime rate in the United States is much lower today than it was in the mid-1980s, notwithstanding the proliferation of cinematic gun play that worries Anderson.

“This research suggests violent media merits our attention,” Anderson concludes. “Exhausting as it may be for parents, being overbearing would seem to pay off.” No one who disagrees with that sentiment will change his mind after reading this study.