A new study published in JAMA Network Open has captured national headlines with its finding that approximately 1 in 15 adults, or about 7%, have been on the scene of a mass shooting. Other startling findings include 2% of the respondents claiming to have not only been present at a mass shooting but injured at one, and over half of those surveyed claimed it happened within the last decade.

But is this finding plausible? Mother Jones’ open-source database of mass shootings defines mass shootings in a very sensible manner which generally aligns with FBI figures: “Our research [focuses] on indiscriminate rampages in public places resulting in four or more victims killed by the attacker. We exclude shootings stemming from more conventionally motivated crimes such as armed robbery or gang violence.” (emphasis added)

While shootings motivated by armed robbery or gang violence are worthy of study, gang violence is not what comes to mind when the media runs headlines with the words “mass shooting.” What comes to mind are socially estranged individuals whose purpose is unmitigated chaos—not protecting gang turf. The underlying causes of gang violence and robbery are totally different than the causes of mass shootings and require explicit differentiation.

Was the JAMA study looking at “mass shootings” as we commonly envision it? The answer is: almost certainly not. Mother Jones has identified 152 mass shooting incidents using their criteria. The database reports a total of 1647 injuries between 1982-2024 as of March 23, 2025. This equates to approximately 38.3 injuries per year. Using today’s population, that amounts to approximately 0.000634% of the population that has been injured at a mass shooting. The study’s conclusion that 2% of the population has been injured at a mass shooting is over 3,100 times greater!

Further, in order for 1 in 15 adults to have been present at a mass shooting as defined in the sensible manner by Mother Jones, there would have to be approximately 114,000 individuals (on average) present at each mass shooting! Even doubling the number of mass shootings to account for the fact that Mother Jones data ends in 1982 would still lead to an impossibly high 57,000 persons present at each shooting. It is clear that the JAMA study was not looking at mass shootings as is typically envisioned by the American people.

What, then, was the study looking at? Because there is no official definition of the mass shooting, the authors had a lot of leeway to define “mass shooting” in a way which could artificially inflate or deflate the numbers at their choosing. The authors simply defined mass shootings as “gun-related crimes where 4 or more people are shot in a public space, such as a school, shopping mall, workplace, or place of worship”—with no gang violence or armed robbery exclusion. While this is an entirely defensible choice, the media has irresponsibly described this as “mass shooting” without the necessary caveat that this includes gang violence and other forms of public shootings the public does envision when one says “mass shooting.”

The data within the paper suggest the findings are driven by gang violence and other more commonplace forms of violence than the relatively rare type of mass shooting that typically grabs media headlines. According to the progressive Violence Policy Center, Asians are the least likely to be victims of homicide, Blacks are the most likely, and Whites are in between. Given that the vast majority of mass shooters knew at least some of their victims, according to a National Institute of Justice report, and mass shootings are committed at about the rate we would expect given the proportion of the population, Blacks, Asians, and Whites ought to have been victims of mass shootings at similar rates to one another if the study was measuring what we typically consider mass shootings.

In the study’s Table 3, it is revealed that Blacks are 87% more likely to report having been present at a mass shooting than Whites and Asians are 64% less likely to report having been present at a mass shooting than Whites.

Not only that, but a third of the respondents claim the shooting occurred in a neighborhood. According to the NIJ report cited above, only 8% of mass shootings occur in residential settings.

These two facts are exactly what we would expect if the study was picking up general violence trends and not violence specific to what we envision as mass shootings.

There are other problems with the analysis. By defining mass shootings explicitly to respondents with locations—”such as a school, shopping mall, workplace, or place of worship”—the results on the location of the shooting are almost certainly tainted with this priming. Further, asking people to recall if four or more people were present at a shooting over their whole lifetime is also unlikely to return an accurate accounting as to whether one was present at a mass shooting.

The most stubborn fact in the paper is that for one in fifteen adults to have been present at a mass shooting, approximately 17.33 million adults would have had to have been present at a mass shooting at least once in their lifetimes. Over half of the respondents claim the shooting occurred at a location where it would be safe to assume the number of individuals present would not be extraordinarily high (neighborhood, bar, workplace, park). A possible estimate for the average number of people present at each of these shootings and survived would be 50—usually less, but sometimes more, especially for those who claim to have been at a shooting at a concert (11%) or shopping outlet (11.5%). The estimate of 50 is almost certainly on the high end (which is what the authors of the JAMA paper need for their results to be plausible). Assuming this, what is the number of mass shootings that had to have occurred in the United States for the paper’s estimate to check out? We can test that using the following equation:

17,333,333 (6.6% of US Adults, or 1-in-15) ÷ X (number of shootings) = 50 (number present and survived)

What is the solution? Well, X = 346,666.66. In other words, assuming 50 people were present at each shooting, the study’s math requires there to have been 346,000 mass shootings in the United States. Let’s triple it to be even more charitable to the authors. Assuming an average of 150 people were present and survived at each mass shooting leads us to conclude there were 115,555 mass shootings in the United States.

Is that possible? I leave it to the reader to determine whether this study’s findings are mathematically plausible and whether or not this should have passed the alleged rigor of peer review.

WHAT TO READ NEXT

WHAT TO READ NEXT