|On Penn’s campus there are organized groups for students who identify as women, Asian, black, queer, and Latino, among others. These groups serve a variety of purposes, and one is to provide a sense of community for types of students who have not traditionally been at home at elite institutions like an Ivy League university. When Lareau considers this list she sees one group missing: an organization for students with lower class backgrounds. “White working class kids who come to college often flounder and they have whole sets of issues as first- generation college students,” she says. “But it’s very hard for those white students to find each other.”
Social class is not a readily recognized category in part because it’s hard to say what exactly it is. Sociologists rely typically on some combination of income, education, wealth, and occupation to measure class position, but there’s no rule for how to combine them. That makes it hard to compare the high-school graduate who owns a painting business and clears a hundred grand a year with the librarian with an advanced degree earning half that. Karl Marx is credited, along with Max Weber, with codifying the idea of social class, but one imagines that when he toured Europe in the 1840s it was easier to see where everyone stood than it is in America today, where forces like immigration, mobility, and the overall high standard of living scramble the social picture.
Given the churn of American society, it’s common for sociologists to talk about class as a continuum instead of a set of strict categories. “I think of social class as more of an opportunity zone that varies by social location,” says Furstenberg, “So I’m not sure whether there are class boundaries or whether it’s a gradient that extends all the way up.”
This is precisely the view Lareau disagrees with. Where sociologists like Furstenberg see a continuum, she sees a few bright lines that separate large portions of the population from each other. Lareau is one of the leading contemporary interpreters of the 20th-century French sociologist Pierre Bourdieu, who is considered one of the foundational voices in social-class theory along with Marx and Weber. Bourdieu argued that a social class is defined by its “habitus,” or its own particular ways of thinking and behaving. Lareau calls this idea “class-specific dispositions” and she argues that when you see people’s lives up close, as she did in Unequal Childhoods, it becomes clearer that “the observable differences in how people act can be meaningfully and faithfully grouped into categories without violating the complexity of daily life.”
The reason many sociologists don’t recognize class distinctions, Lareau says, is because they rely on research methods that aren’t nuanced enough to see them. “You have to take the data you’re given, and the measures of social class on surveys are pretty skimpy, pretty inadequate in my opinion,” Lareau says. “You have to deal with the gradient because that’s the way the data comes.” Most quantitative data in sociology comes from national surveys like the National Survey of Family Growth, which is conducted annually by the Centers for Disease Control and gathers data from tens of thousands of Americans on topics like infertility, pregnancy, contraception, and divorce. The questions that can be asked on these surveys need to be easy to answer, they need to produce data that can be easily categorized, and they need to allow for apples-to-apples comparisons across survey respondents, all of which are necessary constraints but which effectively limit the type of insight that surveys can produce.
A common survey question relevant to the types of differences Lareau studies is, “How many books do you have in your home?” It’s a straightforward question that lends itself to statistical analysis, and it reveals a linear relationship consistent with the idea of class as a gradient: More highly educated people have more books in their homes, but there’s nothing categorical that separates families at any particular point along the line. But Lareau argues that if surveys included more substantive questions that are harder to analyze statistically—like “Describe the techniques you use when reading to your child”—social class divisions would emerge more clearly.
The limited perceptiveness of quantitative analysis is one reason that Lareau became an ethnographer. At Berkeley in the 1970s she met the acclaimed urban ethnographer John Ogbu, who used participant-observation research to conclude that internalized low expectations were a primary driver of underachievement among black youths. “Ethnography as a research tool allows one to bring people to life,” Lareau says. “I remember he [Ogbu] told me you should feel like we’re right on your shoulder, looking over your shoulder. And I think that’s partly what ethnography does, it can show people’s everyday lives.”
Because of the differences in how information is collected, ethnographic conclusions can be hard to assess alongside quantitative research. In order to gain widespread acceptance, quantitative research needs to be reproducible and generalizable: Other scholars need to be able to run the numbers and say, “I find what you find,” and there needs to be good reason to think that what’s true for a group of survey respondents holds for the population as a whole. Social scientists evaluate research against these criteria because conclusions that hold up against them are, generally speaking, more likely to be true than conclusions that do not.
By this standard, ethnography is not a particularly accurate way to create knowledge. Lareau says of her observations in Unequal Childhoods, “If somebody had a similar research question and they were for various reasons able to be in those exact sites again, I would think they would find the same results, more or less.” Of course, those exact sites no longer exist as they did almost 20 years ago. And while Lareau’s 12 family sample was large for an ethnography, and diversified by race and gender, it’s impossible for ethnographic research to meet the standards of generalization used by quantitative scholars. Lareau argues, however, that a narrow view of what counts as methodological rigor has limited the way sociologists think about a range of topics, from social class to public education.
In a 2010 essay in the Teacher’s College Record Lareau argued that education research has suffered in the last decade as a result of a push to study student performance using scientific methods that she says don’t translate well to schools. The 2001 No Child Left Behind (NCLB) Act is closely identified with standardized testing of students, but it also laid out very specific methods that researchers need to use in order to qualify for federal funding. Most notably, NCLB privileged randomized control trials—in which one group of students receives an intervention like after-school tutoring while a control group of students does not—as what it termed the “gold standard” of education research.
Randomized control trials are the best way to produce accurate results in a laboratory, but Lareau argues they don’t work well in settings where researchers cannot control all the variables. “The kind of clean experimental manipulation of ‘conditions’ or ‘treatments’ called for in experimental research is hard to do in real-world settings outside an experimental laboratory,” she wrote in the Teacher’s College Record. “Nowhere are these problems more apparent than in the implementation of randomized-controlled trials in the ‘naturalistic’ settings of schools.”
Lareau cites an experiment in the Chicago public schools as an example of how randomized control trials can go wrong. The experiment called for a group of schools to pilot a program for at-risk youth while a control group continued its pre-existing practices. As it happened, though, several principals in the test group failed to implement the program while several in the control group went ahead with the program anyway because they thought it would help their students. Beyond that basic confusion, other significant disruptions took place which limited the explanatory power of the study. These included “frequent changes in principal leadership, high levels of teacher turnover, shifts in administrative policy, and the placement of one-sixth of the schools in the city on probation.”
Lareau argues that social scientists make a mistake when they rely too heavily on methods that are vulnerable to these types of confounding factors. “There needs to be a realistic and critical assessment of the limits of randomized controlled trials and the relatively narrow forms of knowledge that can be gained from their use,” she wrote. She argues that as a field sociology would better understand the social mechanisms that influence people’s lives if it lent greater legitimacy to the kind of knowledge that is produced through “[s]mall, intensive, non-random case studies” of the sort she undertook in Unequal Childhoods.
Download this article (PDF)
page 1 | 2 | 3 | 4
| ©2011 The Pennsylvania
Last modified 8/26/11