The invitation to convert to Islam comes frequently if you travel in the Muslim world and are not visibly Muslim. Parts of the United States are similar—Have you accepted Jesus as your lord and savior?—although in countries such as Pakistan or Bangladesh or Sudan, the question arrives with even less preamble than its Christian equivalent in the Bible Belt. I have heard the entreaty from strangers on buses, and from shopkeepers craning their neck out their window. It is nearly always friendly, and “no” is an acceptable answer but sometimes elicits dismay and a short lecture. I take no position here on how you should answer, but if your sole purpose is to end the missionary encounter without hurting your host’s feelings, I suggest explaining that your parents would be devastated if you changed religions. Everyone understands the duty to be a good son or daughter, and in most cases you will be left alone.

For many who study jihadism, the most pressing question about conversion is not how to avoid it politely, but why those who do convert tend to join jihadist groups at a disproportionate rate—something like three to four times that of non-convert Muslims (with substantial variation, depending on the country). Vanishingly few converts turn to jihadism, but enough jihadists are converts to invite speculation about the connection. The Islamic State once liked to put converts in front of its cameras, but even off-camera the converts seem to have been overrepresented. Among ISIS-friendly Muslims in Europe, I often met fair-haired recent converts.

[Read: The ‘softer’ side of jihadists]

I used to assume, along with many others, that converts are more zealous in general. After all, they have gone out of their way to join a new faith, whereas non-converts are just born that way. But recent research suggests otherwise. A paper just published by Daniel W. Snook, Lee Branum-Martin, and John G. Horgan examines the religiosity of American Muslim converts and finds that they tend to be less religious than their non-convert peers. Converts, they write, “believe in Islamic tenets less strongly and struggle more with the Islamic faith than” Muslims raised in Islam. European converts were not included in the study, and could well behave differently. I note, too, that “struggling” with one’s faith does not obviously signal lack of religion, as the authors purport. Religious fanatics also struggle.

But let’s say the authors are broadly right, and “the popular notion that converts are more eager, committed, and even fanatical than their non-convert brethren, although anecdotally compelling, may prove to be nothing more than confirmation bias.” The authors do not dispute that converts are overrepresented among the gruesome minority of Muslims drawn to terrorism. So what’s going on? I have a few theories.

Some people convert to please others—to fit in with friends, to satisfy the demand of a spouse’s family. Others convert despite the displeasure of others. Think back to the strategy for getting the recruiter off your back. That strategy works because of a common understanding of filial piety, and a desire to avoid angering or disappointing family. To break the news to your mom and dad that you will have no need for the family Bible when they die, to have a weepy phone call with them and announce that you are changing your name to Muhammad Abdurrahman—these are transgressive acts requiring a steely emotional constitution. Sorry, Mom, but not sorry. (I have even seen cases of people who seem to have converted precisely in order to spite others.)

Snook and his co-authors could not check to see whether the converts to Islam tended in these opposite directions—eager to please in some cases, and in others willing to upset and disappoint those close to them over a matter of principle. Among the converts I have observed who are associated with the Islamic State, disagreeability is a way of life. They are absolutely willing to say things they know I will consider repulsive, if they believe them to be true. Many outsiders wondered why the Islamic State started its reign of terror by reviving the Islamic practice of sex slavery, rather than focusing first on much more attractive Islamic practices such as charity and kindness, then easing into the nasty stuff. I believe this is the answer: ISIS wanted true believers, but not just any true believers; they wanted the most disagreeable and transgressive ones first. Mean people have obvious value on the battlefield. If converts are more likely to be disagreeable than born Muslims, then we’ll see more converts among jihadists, even if converts are on average not more religious.

Another dynamic of conversion may help solve the puzzle. In welcoming converts, born believers naturally assume that the newcomers need guidance. “You will be a baby Muslim,” one missionary told me when seeking my conversion, “and we will help you learn to walk.” But newcomers can be intelligent and in many cases quite headstrong, and this condescension can be grating, particularly if the convert has devoted himself to the diligent study of his new religion and has outdone most native believers in religious knowledge.  

For extremely disagreeable converts, especially those who have gained real understanding of the religion, condescension may sound like a challenge—and some may react by overcompensating and adopting extreme forms of the religion. This happens all the time: the Catholic convert who insists on a Tridentine mass; the Buddhist convert who learns ancient languages and dreams of taking saffron robes. Religious knowledge is not one of the variables that the researchers measured when assessing religiousness. And although the researchers found that converts are less religious, many are still religious. The characteristics of that subset may be what matters in understanding jihadist converts.

[Read: Why jihadists loved America in the 1980s]

The difficult question is where these theories would show up in data. In the general population, only a tiny group is inclined to learn enough classical Arabic to read and understand the Quran (a decade-long intellectual journey, unless done full-time)—and one has to achieve some level of virtuosity to feel the sting of condescension from others less accomplished. The paper surveyed only 177 converts, out of 356 adult Muslims, and I wonder how many of them were several standard deviations more ornery or bookish than the average person. I suspect the researchers would have to find many, many more converts to find substantial numbers of people with these rare traits. The next study should include more screening questions. Are you a convert? Are you also an asshole?