The comprehensive survey of over 1,000 Muslim-Americans released this week by the Pew Research Center was supposed to be a harbinger of good news, as evidenced by its title, "Muslim Americans: Middle Class and Mostly Mainstream". And in many ways, it has positive things to report: most Muslim Americans buy into American ideals of hard work and opportunity, have many non-Muslim friends, are relatively educated and well off (only 2% are low income), and report being "happy" or "very happy" with their lives.

The survey also showed that Muslim-Americans views towards Israel are in line with other Americans (most believe that Israeli and Palestinian rights can be reconciled) and that they categorically reject extremism among Muslims. "What this survey shows is that Muslim Americans are largely assimilated, happy with their lives and moderate - mostly in contrast to Muslims in western Europe," said Andrew Kohut, head of the Pew Research Center. "They also reject Islamic extremism to a much greater extent than Muslim populations elsewhere in the world."

This might be the case, but you wouldn't know it from reading the headlines covering the study, most of which are focusing on one troubling statistic: 8% of US Muslims - and 15% of US Muslims under 30 - believe that suicide bombings can be often or sometimes justified in the defense of Islam. With an estimate of 2.35 million Muslims in the US, this statistic has predictably caused some degree of alarm over the 140,000 or so Muslims that fall into this troublesome camp. "Jihad in America?" reads one headline, with other similar articles attracting angry comments. "It is a hair-raising number," admits Radwan Masmoudi, president of the Washington-based Center for the Study of Islam and Democracy. Indeed it is.

But something is missing from this poll that any scientist would understand - a "control." That is, one needs to ask non-Muslim Americans the same questions about terrorism to see where the answers deviate. Fortunately, one such poll with an identical question was released a few months ago and, though it didn't result in any headlines, the deviation is remarkable - and unexpected. When asked if "bombing and other attacks intentionally aimed at civilians" are "often or sometimes justified," 24% of all Americans agreed - three times the 8% of US Muslims who share that view. And scarcely half as many non-Muslims (46%) were found to oppose suicide bombings as all US Muslims in the Pew poll (80%). Similarly, a UK poll last year cited support for suicide bombing among Muslims (10%) and non-Muslims (7%) that is statistically equal. The Pew finding that 47% of US Muslims consider themselves Muslim first - rather than American - needs to be compared to a similar poll that showed 42% of Christians and 62% of white Evangelicals identifying themselves primarily by their religion as opposed to their being American.

Using the alarmist logic that is currently being applied to US Muslims, the numbers would mean 72 million Americans are walking time bombs (or in support of them). Fair? Of course not. But the lack of a "control" demonstrates the inherent flaw in interpreting answers to questions like these - and both Pew researchers and rushed reporters should have known better.

Glenn Greenwald at Salon has more on this.

Speaking of "control," the list of those polled was "constructed from a commercial database of households where someone in the household has a name commonly found among Muslims."

That might be fine for marketing purposes to selling goods to potential Muslim customers, but to me it sounds a bit suspect. I usually have total respect for Pew researchers, but this fly in the ointment of their due diligence in this case should raise their margin of error (at 5%) to almost unacceptable levels.

For example, many Iranians have "Muslim" sounding names, but have a bad taste of Islam in their mouth after fleeing the Iranian revolution. Similar reasons could be said for Arab Christians.

Why couldn't they get a hold of a phone list from, say, ISNA or MPAC or CAIR or, even better, John Ashcroft?

Only 2% with low income??! That does not sound right unless they excluded alot of types of Muslims: refugees, inner-city residents, rural residents and so on. Every community I lived in had immigrant taxi drivers living 4 to a room, for example and plenty of people who were low income. Thier results must be skewed in some significant way as to whom were selected for the survey.

>> a phone list from, say, ISNA or MPAC or CAIR

That's a just as bad bias as the way they did it because it self-selects members of only those organizations who total less than 3% of the Muslim community in the US. So such a survey could only represent the opinions of ISNA-ites, CAIRies, and MPACers who may not represent the aggregate opinions of the whole.

Anyway, good on Shahed for pointing out the need for a control. All opinions can be found in any population, even if they are statistically small, so the only way to derive knowledge from that information is to compare it with something else.

A few points in defense of the Pew Report researchers:

Controls
The goal of the Pew Report was to develop a snapshot portrait of Muslim-Americans, in their words, "to measure rigorously the demographics, attitudes and experiences of Muslim Americans." It was not to prove that Muslim-Americans or more likely to believe certain things than other Americans. In such a case, control populations are like a garnish. They are nice to have, but not essential. To have developed a control population for this study would have taken significantly more time and effort than what their goal required.

Where possible, the actual report contextualizes the data on Muslim-Americans by providing answers to similar (often identical) questions from Americans broadly, sub-groups of Americans (e.g. Christians and African-Americans), other nationalities and Muslims of other nationalities. Among these comparisons is one comparing the extent to which Christians and Muslims see themselves as of their faith first rather than of their nation-states first. In many cases, they are using data from other Pew sources, because they can attest to the similarity of the questions and the conditions under which they were asked. (Believe me, not all surveys are equal.) Typically, serious researchers will not use other questions for base comparisons that are not similarly framed and administered.

...

By using only "Muslim" sounding names culled from official sources, they automatically excluded converts, a number of whom do not necessarily list our names in phone books that sound Arabic. So, they drew a portrait of exacty which so-called Muslim-Americans??

...

Population Sampling (Selection)
Getting an adequate national sample of a minority population is quite difficult. This is the case for African-Americans who are about 12% of the population and must be doubly so for Muslim-Americans who are less than half that proportion of the population. At least in the case of African-Americans, one has a head-start because US Census data includes race. It does not include religious identification. The Pew Report pulled its sample from three sources: a geographically-stratified random cold-call sample, a commercial database from Experian (a national credit bureau) of persons with first names and surnames (deemed likely to belong to Muslims) and persons who completed previous Pew (and other research bureau) surveys who had already self-identified as Muslim. In each case, the respondent was asked if her or she was Muslim. (Persons who were Muslim had their data included and those who were not did not.)

From the report:
The initial phase of the questionnaire included neutral or innocuous questions about satisfaction with the community, personal happiness, and personal characteristics such as home ownership, entrepreneurship, and newspaper subscription. After these items, respondents were asked about their religious affiliation, choosing from a list that included major Western traditions such as Protestantism and Catholicism but also non-Western traditions of Buddhism, Hinduism, and Islam. Respondents who identified as Muslim proceeded to the substantive portion of the questionnaire, and those who were not Muslim were asked if anyone in the household practiced a different religion; in 58 households where a non-Muslim answered the phone, it was determined that there was a Muslim living in the household, and 52 of these subsequently yielded a completed interview.

...

The report was not designed to understand Muslims who just had good experiences with the religion or the practice of the religion, so it did not exclude based on that criterion. Second, getting a list from one of the organizations above would just introduce another set of biases, that is people who have had contact with, who support or who somehow ended up on the mailing list of one of those groups. It would be equivalent to me sampling African-Americans by getting a list from the NAACP or The Urban League.

I do have some minor concerns about the sampling, but am not sure how I would have fairly resolved them.

First, they included persons who said they were Muslim and apparently listed their affiliation as the Nation of Islam.

From the report:
Overall, 20% of U.S. Muslims are native-born African Americans, nearly half of whom (48%) identify as Sunni. Another third (34%) of native-born African Americans say they are just a Muslim, and 15% have another affiliation, including Shia and the Nation of Islam.

I guess excluding them would have involved the researchers in a debate about who can call themselves Muslim.

Second, their phone numbers were restricted to landline numbers--an increasing liability in the world of polling given the ubiquity of cell phones. Historically, such restrictions were believed to reduce the number of very poor respondents. I'm not sure how true that still is. Also, as they admit in the report, the Experian data seems to skew towards better-educated and higher-income respondents. (The Experian database probably accounted for about 40% of the respondents.)

In summary, I think the Pew research team just wanted to add to the scant data about who Muslim-Americans are. (There's very little national quantitative data on this subject.) The report seems to me to be as thoughtful as exposition of poll-data can be and fair to Muslim-Americans. If one wants to blame someone, try the news media, but please don't blame the Pew researchers.

If one wants to blame someone, try the news media, but please don't blame the Pew researchers

What I blame the Pew researchers for is divorcing themselves from the way that the data, however accurate, is being used in the media. We're not talking about personal tastes or political preferences, for which poll numbers wouldn't merit front-page news. We're talking about support of suicide bombings, which is particularly sensitive stuff. It really should have been featured alongside the views of non-Muslims on the same subject to neutralize the potential misuse of the numbers.

Shahed,

You've made an excellent point about the need to put the findings on young Muslim support for suicide bombing in context. Andrew Kohut, president of the Pew Research Center, noted at the press conference that young people of all backgrounds, when polled, are more likely to support violent action than older people.

Andrea
My posting on the Pew study: http://religionwriter.com/?p=69

There are a few problems with this poll, as some people have already alluded to. My main question is how they came up with only 2.35 million Muslims in the U.S? Previous studies have estimated the number between 7 and 10 million. Even in Canada, with a population of approx. 30 million people, have approx. 1 million Muslims, so it doesnt make any sense to me, how in the U.S. with 300 million people, only 2.35 million Muslims?

Also it is obvious that most of the Muslims selected for the poll are from Immigrant communities, because again the numbers do not jive with previous reports which state that almost 30-40% of American Muslims are of African-American descent. Also most immigrant Muslims are also more well off and also consider themselves "happier" than majority of African-American and Latino Muslims, or those who converted.

Great points, Mr.Amanullah. I have also read and referenced the CSM article on American approval of attacks on civilians. People can interpret what they want from this study but it makes me wonder how the questions were worded. Americans are masters of media manupulation and propaganda so I can see how they can run with this and distort the results to suit their own agenda.

In psychology of mass media there is a saying:

"Whoever controls the questions....controls the answers"

Irfan
[url=http://www.MediaAndIslam.com]http://www.MediaAndIslam.com[/url]

Mr. Amanullah included a link to the report. A copy of the survey instrument as well as the report and methodology are there. I suggest that it should be read or at least skimmed because it would answer some of the questions here.

Aslaamun alaykum all,

Any time you generate numbers on polarising topics, you're going to get people who use them for cynical purposes. The Cosco Coulter, Debbie Schlussel, predictably went lateral over the poll, as did Mark Steyn and a host of others. That's going to happen no matter how you contextualise the results. People love playing with numbers.

Now, a bit of context doesn't hurt, but from the perspective of concerned Muslim outreach, it doesn't necessarily help either. What difference would it make if a greater proportion of non-Muslim youth think suicide bombings can be justifiable than Muslim youth? To me, it's an interesting window dressing, but not my concern. What does concern me is that a number greater than zero of Muslim youth believe that, which is alarming no matter the context, and compareable results from other communities doesn't change that.

Thanks again for the article.

--A

What does concern me is that a number greater than zero of Muslim youth believe that, which is alarming no matter the context, and compareable results from other communities doesn't change that.

Good point - I hope that people realize that I did not mean to downplay the obvious problem that this is. (See my last article.)

I don't think contextualising the Pew research results downplays them; it helps show that these aren't particularly Muslim problems. But the problems do have particularly Muslim solutions, and I think our communities would do well not to lose sight of that. Don't get me wrong: it's valuable to get to the true nature of the problem, and the contexts you provide above help to do that--immensely, in fact.

" a similar poll that showed 42% of Christians and 62% of white Evangelicals identifying themselves primarily by their religion as opposed to their being American. Using the alarmist logic that is currently being applied to US Muslims, the numbers would mean 72 million Americans are walking time bombs (or in support of them). Fair? Of course not. But the lack of a "control" demonstrates the inherent flaw in interpreting answers to questions like these"

That charge is a red herring, because there is no Christian equivalent to Al-Q and Salafism in the U.S. As there are no violent extremist Christian groups in the U.S., the end points are different, so to speak.

While the survey has been represented in the media as proof of moderation among American Muslims, the actual results should yield the opposite conclusion. If, as the Pew study estimates, there are 2.35 million Muslims in America, that means there are a substantial number of people in the U.S. who think suicide bombing is sometimes justified. Similarly, if 5% of American Muslims support al Qaeda, that's more than 100,000 people.

As there are no violent extremist Christian groups in the U.S....

The problem with making such an emphatic and self-sure statement like that is that it's simple to prove emphatically and assuredly false. There are plenty of violent extremist Christian groups in the US. But that doesn't matter. A fellow merely has to point to one, like the Aryan Nations or the Phineas Priesthood or any other of the Christian Identity movement groups, for your statement to be untrue. In the future, try using weasel-words, like "relatively fewer violent extremist Christian groups," to mitigate against that.

Hope that helps,
A

Abdiel, those groups are not examples of Christian extremism, but of racism. Their ideology is not founded in the Christian religion, nor can they immerse themselves in a community of passive supporters. Furthermore, the activities of such "extremists" - the FBI's highest terrorist priority in the 90's - are far more closely monitored than Islamic radicals, and are not nearly the same level of threat.

These are guys who might kill somebody either Jewish or Muslim, but for the moment anyway, I'm not worried about them as an existential threat. And if I'm not worried, why should you be?

You can read more about these White Supremacist groups at the website of the Anti-Defamation League, which works on countering them. The ADL takes donations from everybody, regardless of religion. Even CAIR has had nice things to say about them!