Headscarves
In this study, subjects were randomly assigned to view a picture of a woman or a picture of this same woman wearing a headscarf in the style of some Islamic women. Here are the two pictures:
The headscarf had some dramatic effects:
1) The covered woman was perceived as more “traditional” and, in personality terms, less “warm.” She is also described as living a more insular life. Here are the percentages describing the two women in a variety of ways. (First percentage: uncovered; second percentage: covered.)
Age 36 or older: 15% vs. 30%
Marital status is single: 59% vs. 25%
Assuming woman is married, she is not working outside the home: 12% vs. 47%
A good mother: 33% vs. 45%
A devoted wife: 26% vs. 51%
Lively: 60% vs. 40%
Has a sense of humor: 61% vs. 37%
Always looks on the bright side: 60% vs. 43%
Might be the life of the party: 26% vs. 6%
Sticks to a tight circle of people: 24% vs. 43%
Keeps to herself: 8% vs. 22%
Strict: 2% vs. 23%
2) The covered woman was perceived as wealthier. (People imagine a wife of the “rich sheikh” stereotype, perhaps.) An equal number of subjects considered the two women “stylish.” Slightly more considered the covered woman “beautiful” (27%) than did the uncovered woman (16%).
3) The vast majority of respondents thought the uncovered woman was “an American” (82%). The vast majority of subjects thought the covered woman was “a Middle-Eastern person” (78%) and also Muslim (87%).
4) Subjects displayed considerably more aversion to the covered woman. Specifically, they were less likely to want to live near her. While 89% said that they would like the uncovered woman as their next-door neighbor or in their neighborhood, only 62% said that about the covered woman. One-fifth (19%) actually said they wanted her to live “outside of the US.”
What is the meaning of this hypothetical exercise? These pictures of unfamiliar people encourage subjects to engage in a common cognitive process: categorization. They subconsciously place the woman in a group and then impute to this woman the perceived characteristics of that group. These group characteristics are also known as stereotypes. Obviously, the sense that this woman lives a traditional life, is not lively or warm, and keeps to herself is quite in line with common stereotypes of Muslims.
This kind of experiment, despite its artificiality, actually has a great deal to tell us about the real world. Many Americans will see a woman wearing a headscarf only in passing, without any substantive interaction. Given only a brief “snapshot” of the person, people will probably then engage in the same kind of categorization process that these experimental subjects engaged in.
[Hat tip to a GW student, Joseph Essex.]
Comments
What is more, these stereotypes are sometimes self-fulfilling as shown in classic study by Word, Zanna and Cooper (1974).
Posted by: Jacob Sohlberg | January 24, 2008 06:42 PM
Some of the characterizations (not all) seem like statistically rational guesses: "sticks to a tight circle of people," "strict." Others are not: age, marital status. And some are just troubling: a covered woman is more attractive? Really?
Posted by: anonymous | January 28, 2008 10:34 PM
While I suspect there is plenty of underlying U.S. bias toward Islam in these results, the overall implications seem to be more about lifestyle than culture. I expect any young woman who expresses her devout faith so clearly might be met with similar responses. I expect that real statistics about devout women of any faith would support the idea that she may be more likely to marry early, raise children more strictly, be a devoted wife, and be more likely to be a stay-at-home mother.
When it comes to personality issues like being the "life of the party" or "sticks to a tight circle", those attributes are probably disproportionately true of most women who wear head scarves in the U.S., largely because they are often excluded as the "other" in many circumstances. Clearly the numbers in this poll are dramatic, but that doesn't mean there isn't some underlying reason for participants to perceive her that way. She may be a laugh riot, but none of the WASP neighbors would have any idea. And if she grew up treated differently, you can bet that sense of humor was stifled a bit, or at least shared with fewer people.
And as for beauty, analysis of hair style is subjective, but analysis of basic facial features is primal. When you focus on the face, beauty is simply beautiful, and hair only complicates that.
Just my .02.
Posted by: Rob | January 29, 2008 11:46 PM
That's so funny, because I conducted a study where there were two pictures of a man - one was wearing a wedding ring and the other was not. 100% of respondents regarded the second man as being married, whereas 0% of respondents regarded the first man as being married. We really are a close-minded species.
Posted by: Anonymous | January 31, 2008 12:21 PM
Anonymous, your wedding rings example isn't apt, and the head covering study isn't that obvious. Your example would be apt if you found the man with the wedding ring was perceived as having very different personality attributes than the unmarried man.
Rob, you could be right about the ways in which people perceive religious orthodoxy generally. The key difference, of course, is that the visual cues of Muslim orthodoxy are more obvious than those of Christian orthodoxy, making the likelihood of categorization and stereotyping greater.
Posted by: John Sides | January 31, 2008 02:38 PM
I have doubts that such a narrow experiment produces reliable results. First, there is no control. As such, it seems obvious that many people answering questions will perceive what is being asked and coach their responses. You'd need a wider group of examples than just two. Otherwise, you're just going to get stereotypical responses. Basically all this experiment shows us is that people know how they're supposed to respond to such visual cues. It doesn't really tell us what people really think.
Posted by: ern | January 31, 2008 03:08 PM
Ern, I don't know what you mean about the lack of control. What would be an appropriate "control"? And what other examples would you suggest?
Second, if the people who were randomly assigned to the woman with a headscarf realized what was going on, then likely they provided more "tolerant" responses, and thus the experiment may *underestimate* people's less favorable preconceptions of Muslims.
Moreover, "stereotypical responses" are precisely the point of the results: people use stereotypes to attribute qualities to the woman. That may not be surprising, but findings don't have to be surprising to be important.
Posted by: John Sides | January 31, 2008 08:45 PM