We are almost always interested in knowing how the public thinks and feels about various issues -- global warming, race relations, the fairness of rising income inequalities, and the acceptability of same-sex marriage, for example. The public is composed of millions of individuals, and the population can be segmented in a variety of relevant ways -- gender, age, race, region, political affiliation, and many other cleavages. So we might want to know how teenagers think about alcohol, and how these attitudes have changed over time, or how attitudes towards the equality of women have evolved since 1960.
What research tools are available to us to investigate public opinion? And how reliable are these tools?
The most immediate answer to this question is survey research. We can formulate a set of survey questions, select a population of respondents, and tabulate the distribution of responses. And if we do this on several occasions over time, we can make some inferences about changes in attitudes over time. There are innumerable examples of these kinds of studies, from the GSS to the euro barometer to the Pugh organization's frequent polling data. And we seem to learn some important things from studies like these concerning the distribution of attitudes across time and space. Spaniards are less concerned about fair trade produce than Swedes, young people have become more accepting of single-sex marriage, people over 65 are more sympathetic to Tea Party values than people in their thirties.
Another research approach takes a more qualitative approach. Researchers sometimes use open-ended interviews and focus groups to learn more directly how various groups and individuals think about certain topics. We may learn more from such studies than we can learn from a mass survey -- for example, the interviewer may gain a better understanding of the reasoning that individuals use to reach their beliefs. A survey question may ask whether a consumer is willing to pay 10% more for fair trade bananas, whereas a series of interviews may determine that the "no's" break into a group who don't have the discretionary money and another group who oppose fair trade pricing on ideological grounds.
There are more indirect methods for studying public opinion as well. We might examine the comments that are submitted to newspapers on topics of interest and try to quantify over time the "temperature" of those comments -- more intolerant, more angry, more reflective. Likewise, we might attempt to quantify the streams of social media -- Twitter, Facebook, tumblr -- with an eye towards testing the attitudes and emotions of various segments of the public. Consider he vitriol that exploded on twitter following the selection of Miss New York as Miss America a few weeks ago.
Attitudes towards race in America are especially interesting to me. How have Americans changed in the ways they think about race? Have Americans become less racially intolerant since the civil rights movement decade? Survey data seems to give a qualified "yes" to this question. Answers to survey questions that explicitly probe the individual's level of racial antagonism seem to support the notion that on average, antagonism has declined. But scholars in race studies such as Tyrone Forman and Eduardo Bonilla-Silva argue that the survey results are misleading (link). They use a methodology of extended interviews, along with a rigorous way of interpreting the results, which suggests that there is a widening divergence between the responses American college students give to surveys probing their racial attitudes and the values and beliefs that are revealed through extensive open-ended interviews.
In consonance with this new structure, various analysts have pointed out that a new racial ideology has emerged that, in contrast to the Jim Crow racism or the ideology of the color line (Johnson, 1943, 1946; Myrdal, 1944), avoids direct racial discourse but effectively safeguards racial privilege (Bobo et al., 1997; Bonilla-Silva and Lewis, 1999; Essed, 1996; Jackman, 1994; Kovel, 1984). That ideology also shapes the very nature and style of contemporary racial dis-cussions. In fact, in the post civil rights era, overt discussions of racial issues have become so taboo that it has become extremely difficult to assess racial attitudes and behavior using conventional research strategies (Myers, 1993; Van Dijk, 1984, 1987, 1997). Although we agree with those who suggest that there has been a normative change in terms of what is appropriate racial discourse and even racial etiquette (Schuman et al., 1988), we disagree with their interpret-ation of its meaning. Whereas they suggest that there is a ‘mixture of progress and resistance, certainty and ambivalence, striking movement and mere surface change' (p. 212), we believe (1) that there has been a rearticulation of the domi-nant racial themes (less overt expression of racial resentment about issues anchored in the Jim Crow era such as strict racial segregation in schools, neigh-borhoods, and social life in general, and more resentment on new issues such as affirmative action, government intervention, and welfare) and (2) that a new way of talking about racial issues in public venues – a new racetalk – has emerged. Nonetheless, the new racial ideology continues to help in the reproduction of White supremacy. (52)
They argue that a new "racetalk" has emerged that makes explicitly racist utterances socially unacceptable; but that the underlying attitudes have not changed as much. And this implies that survey research is likely to misrepresent the degree of change in attitudes that has occurred.
Here is a good example of the kind of analysis Forman and Bonilla-Silva provide for transcripts from the open-ended interviews:
The final case is Eric, a student at a large midwestern university , an example of the students who openly expressed serious reservations about interracial mar-riages (category 6). It is significant to point out that even the three students who stated that they would not enter into these relationships, claimed that there was nothing wrong with interracial relationships per se. Below is the exchange between Eric and our interviewer on this matter.
Eric: Uh . . . (sighs) I would say that I agree with that, I guess. I mean . . . I would say that I really don't have much of a problem with it but when you, ya know, If I were to ask if I had a daughter or something like that, or even one of my sisters, um . . . were to going to get married to a minority or a Black, I . . . I would probably . . . it would probably bother me a little bit just because of what you said . . . Like the children and how it would . . . might do to our family as it is. Um . . . so I mean, just being honest, I guess that's the way I feel about that. Int.: What would, specifically , if you can, is it . . . would it be the children? And, if it's the children, what would be the problem with, um . . . uh . . . adjustment, or Eric: For the children, yeah, I think it would just be . . . I guess, through my experience when I was younger and growing up and just . . . ya know, those kids were different. Ya know, they were, as a kid, I guess you don't think much about why kids are dif-ferent or anything, you just kind of see that they are different and treat them differ-ently . Ya know, because you're not smart enough to think about it, I guess. And the, the other thing is . . . I don't know how it might cause problems within our family if it happened within our family, ya know, just . . . from people's different opinions on some-thing like that. I just don't think it would be a healthy thing for my family. I really can't talk about other people. Int.: But would you feel comfortable with it pretty much? Eric: Yeah. Yeah, that's the way I think, especially , um . . . ya know, grandparents of things like that. Um, right or wrong, I think that's what would happen. (Interview # 248: 10)
Eric used the apparent admission semantic move (“I would say that I agree with that”) in his reply but could not camouflage very well his true feelings (“If I were to ask if I had a daughter or something like that, or even one of my sisters, um . . . were [sic] to going to get married to a minority or a Black, I . . . I would probably . . . it would probably bother me a little bit”). Interestingly , Eric claimed in the interview that he had been romantically interested in an Asian-Indian woman his first year in college. However, that interest “never turned out to be a real big [deal]” (Interview # 248: 9). Despite Eric's fleeting attraction to a person of color, his life was racially segregated: no minority friends and no meaningful interaction with any Black person.
This is an important argument within race studies. But it also serves as an important caution about uncritical reliance on survey research as an indicator of public attitudes and thinking.