Amid the extreme and often emotional reactions around the world to the US government’s suspension of travel and refugees from seven Middle Eastern countries on counterterrorism grounds, it’s often hard to find real facts.

A joint report released this week by the University of Chicago’s Chicago Project on Security and Threats (CPOST) and the Australian Strategic Policy Institute is the first comprehensive analysis of ISIS-related cases to examine the profiles of indictees overall, as well as to identify characteristics associated with each of the three categories of terrorism offences. Our findings challenge conventional stereotypes of terrorists and even the standard profiles of past groups.

Commentary to date on the type of people in the US who support ISIS is typically based on a few high-profile cases and considerable speculation. To understand this new American face of ISIS, the study examined 112 cases of individuals who perpetrated ISIS-related offences, were indicted by the US Justice Department for such offences, or both, between March 2014 and August 2016. The three types of offences are:

1. attacking or conspiring to attack targets in the US
2. travelling or conspiring to travel to join ISIS abroad
3. facilitating others seeking to attack or travel

So what did we find?

First, US ISIS indictees look more like average Americans than is commonly understood. While the image of the ‘typical terrorist’ is that of a young, single male under the age of 25 years, the profile emerging from our research paints a different picture. US ISIS indictees are older—nearly half are over 25—and a notable fraction (11%) are women. In addition, their rates of marriage and higher education are comparable to the US national average, and three-quarters were either students or employed at the time of the offence.

In short, they’re engaged with society and have educational and career opportunities. They aren’t loners operating from the fringes of society. Nevertheless, their opportunities and social relationships didn’t prevent them being radicalised and becoming active supporters of ISIS.

Second, the indictees are truly homegrown. The vast majority are US citizens (83%), and 65% were born in the US. None of them are Syrian refugees. Indeed, only three of the 112 had refugee status at the time of their offences, and two of those had arrived in the US before 1999. Two of the three were from Bosnia and one from Iraq. However, a significant fraction of those born in the US are second-generation Americans, consistent with studies investigating ISIS recruitment in other Western countries, such as France. While data on the families of US indictees is limited, we know that at least 17 were born into Muslim immigrant families, and evidence points to an additional four for whom that’s highly likely (together comprising 29% of the 73 US-born indictees).

Third, many indictees come from outside established Muslim communities. Half (51%) of those who chose to attack in the US are recent converts to Islam, including some who converted less than a year before their arrest. That’s in sharp contrast to the smaller number of converts among those who chose to travel to fight in Syria (19%) or who facilitated attackers and ‘travellers’ or foreign fighters with money and logistical support (10%). Travellers were, on average, the youngest offenders, at 25 years old. Facilitators weren’t only the oldest, averaging 29 years, but also the least likely to be unemployed.

Fourth, ISIS propaganda, and especially videos, played a central role in their radicalisation. 83% of indictees reported watching ISIS videos, including videos of lectures and executions, which ISIS distributes widely on the internet.

Finally, ISIS has been more successful than al-Qaeda in mobilising support in the US, with four times as many people charged annually. Significantly, ISIS indictees are also more likely to be US citizens and recent converts than their al-Qaeda counterparts. ISIS’s comparative success underscores not only the effectiveness of its propaganda strategy but also the centrality of the internet in making the group’s propaganda available to potential supporters across the globe.

Taken together, the increase in ‘born and bred’ Americans supporting ISIS and the very limited number of refugees in our study suggest that limiting or halting immigration from Muslim countries won’t eliminate or even markedly mitigate the threat posed by ISIS to the US. Significantly, in the light of current debates about security threats associated with Syrian refugees from the current conflict, there aren’t cases in our data of a refugee from Syria perpetrating an ISIS-related offence in the US.

That said, this study did not review ongoing cases of terrorism-related investigations in the US that haven’t yet proceeded to a charge. Nor was it in the scope of this study to examine the current or future threat ISIS poses to the US and how the group might seek to act on this.

Because our study doesn’t point to a narrow, easily distinguishable profile, counter-terrorism agencies must focus on limiting access to the tools used to carry out attacks and the propaganda that inspires them as much as on more traditional streams of information to identify emerging threats.

To do so, we need to deepen our understanding of the appeal of this propaganda in order to sever the link between individuals’ sense of disenfranchisement, perceived lack of opportunity and other factors, as well as their support for ISIS.