Some of the protesters who marched through Charlottesville last weekend were described as “alt-right,” a newish term that has been used for everyone from white supremacists to economic populists. But what does it actually mean? The Associated Press recently issued guidelines discouraging journalists from using the term “generically and without definition” since “the term may exist primarily as a public-relations device to make its supporters’ actual beliefs less clear and more acceptable to a broader audience.” Meanwhile, President Trump recently told reporters that some of the protesters in Charlottesville who waved Nazi insignia and chanted anti-Jewish slogans weren’t all nefarious—some “were very fine people.”

A psychology paper put out just last week by Patrick Forscher of the University of Arkansas and Nour Kteily of Northwestern University seeks to answer the question of just what, exactly, it is that the alt-right believes. What differentiates them from the average American?

For the paper, which has not yet been peer-reviewed, Forscher and Kteily recruited 447 self-proclaimed members of the alt-right online and gave them a series of surveys. How did they know these people were really “alt-right?” The individuals responded to questions like, “What are your thoughts when people claim the alt-right is racist?” with statements like:

“If it were not for Europeans, there would be nothing but the third world. Racist really needs defined. Is it racist to not want your community flooded with 3,000 low IQ blacks from the Congo? I would suggest almost everyone would not. It is not racist to want to live among your own ... Through media [the Jews] lie about the Holohoax, and the slave trade. Jews were the slave traders, not Europeans ... many people don't even understand these simple things.”

The researchers compared the responses of the alt-right people to a sample of people who did not identify as alt-right. What they found paints a dark picture of a group that feels white people are disadvantaged. They are eager to take action to boost whites’ standing. What’s more, they appear to view other religious and ethnic groups as subhuman.

Importantly, the study authors did not find that economic anxiety was driving the alt-right’s sentiments, debunking a popular theory in the wake of the 2016 election. “Alt-right supporters were more optimistic about the current and future states of the economy than non-supporters,” they write.

But there were key ways that the alt-right participants differed from the comparison group. The alt-right members trusted “‘alternative’ media” such as Breitbart and Fox more than mainstream outlets. They were much more likely to have a “social-dominance orientation,” or the desire that there be a hierarchy among groups in society.

One can easily guess who they want at the top of this hierarchy. The alt-right participants were more likely to think men, whites, Republicans, and the alt-right themselves were discriminated against, while minorities and women were not. This is in line with past research showing that white supremacists have a victimhood mentality, in which they consider whites to be the real oppressed people of American society.

In this study, the alt-right members were much more likely to be willing to express prejudice, to engage in offensive behavior and harassment, and to oppose Black Lives Matter. And here’s the scariest part. The researchers showed the participants the below scale, which psychologists use to ask people how “evolved” various groups are. A score of zero puts them closer to the ape-like figure on the left, while a 100 is the fully evolved human on the right. It’s a scale, in other words, of dehumanization.

Forscher/Kteily

The alt-right members were much more likely to consider groups they see as their opponents—people like Muslims, Mexicans, blacks, journalists, Democrats, and feminists—to be less evolved than they are. “If we translate the alt-right and non-alt-right ratings into their corresponding ascent silhouettes, this means that our alt-right sample saw religious, national, and political opposition groups as a full silhouette less evolved than the non-alt-right sample,” the authors write.

Vox’s Brian Resnick further breaks down the data here:

On average, they rated Muslims at a 55.4 (again, out of 100), Democrats at 60.4, black people at 64.7, Mexicans at 67.7, journalists at 58.6, Jews at 73, and feminists at 57. These groups appear as subhumans to those taking the survey. And what about white people? They were scored at a noble 91.8. (You can look through all the data here.)

The comparison group, on the other hand, scored all these groups in the 80s or 90s on average. (In science terms, the alt-righters were nearly a full standard deviation more extreme in their responses than the comparison group.)

“If you look at the mean dehumanization scores, they’re about at the level to the degree people in the United States dehumanize ISIS,” Forscher says. “The reason why I find that so astonishing is that we’re engaged in violent conflict with ISIS.”

Forscher and Kteily also found there were two distinct subgroups in their sample of alt-righters. Some were “populists,” who were concerned about government corruption and were less extremist. The more extreme and racist among them, meanwhile, were the “supremacists.” The authors speculate that people who start out as populists might become radicalized into the supremacist camp as they meet more alt-righters.

This study, once it is peer-reviewed, may have broad implications for the fight against hate groups—and for psychology itself. As the authors note, modern psychology studies mostly focus on implicit bias—the internal racism that most people don’t outwardly express. They might be, say, slower to associate “professor” with a picture of an African-American person, but they’re not grabbing torches and heading to rallies. Perhaps psychologists simply thought society had progressed to the point where overt racism is so rare as to be difficult to measure. But this study shows that hundreds of actual, proud racists can be easily recruited online for a study for the low price of $3.

The authors of this paper write that “blatant intergroup bias has by no means disappeared.” It’s something the events in Charlottesville revealed all too vividly last weekend.