Friday, March 7, 2014

Problems with measuring attitudes

Attitudes towards confessional and sexual minorities in the discourse of the media in Latvia, Lithuania and Poland from a historical perspective 

Why do we need to do discourse analysis in order to retrieve social attitudes towards minorities? Why can we not rely on social surveys, questionnaires, interviews? 

There are two reasons why my project does not use these methods: 1) the hypothesis that attitudes are not fixed mental entities, but discursive constructions (see above), 2) problems associated with opinion polling (discussed below). 

Ever since the beginning of the practice of opinion polling, it has been widely criticized: “The history of opinion polling over the past 80 years has constituted a striking attempt to attribute ideas to the public in ways that are discrete and cohesive, descriptive and predictive, illuminating and shaming. The public cannot appeal against misrepresentative claims about its opinions, for opinion polling not only defines such opinions but appears to define the public itself. In short, the public’s scientifically measured presence has come to be regarded as a more legitimate reality than its autonomous attempts to speak for itself. The crowd came to be seen as wholly observable, explicable, and predictable” (Coleman & Ross 2010: 14). 
“We would argue that polls are always discursively situated, constitutive techniques which do not merely capture preexisting opinion, but conjure it into existence. In other words, what political scientists refer to as “public opinion” is in fact what pollsters decide to poll. A sceptical view of the so-called neutral scientificity of opinion surveys accords with our constructivist conception of the public. (…) The search for public opinion is never neutral; like all social techniques, it is prompted by particular intentions which are often left unstated” (p. 15, emphasis mine). 
That public opinion is what pollsters decide to poll is a very important point. People are not asked what they think, they are asked what they think about a specific issue, and they may feel forced to answer even if they have not formed an opinion about it, do not care about it, consider it a private matter not to be discussed with strangers, and so on. When a problem is discussed in an opinion poll, it automatically becomes a matter of public debate, even if there are many other problems that people care about more. 

 There is also the problem of holding unpopular, discriminatory views. “Even in advanced democratic societies with well-functioning public realms, racial, religious, and other group prejudices belong in a category of opinions that are often kept hidden because their public expression would amount to an open breach of the consensus rejecting such views” (Kovács 2005: 270). Racists, for example, may decide to hide or downplay their views in an opinion poll, especially if they consider it a sort of public, not private discourse. The poll may be distorted by latency. “Scholars concerned with the problem distinguish between two forms of latency. They speak of conscious or factual latency where people have no developed opinions about certain issues, and of communicative or functional latency where participants in the communication hide their real opinions” (p. 270). “According to the theory of “the spiral of silence” people in communication try to judge whether open expression of their views would provoke conflict with uncomfortable psychological consequences. Those who feel they hold a minority opinion in society often tend to avoid to say what they really think” (p. 272). 

Another problem with measuring attitudes is that surveys and questionnaires often force respondents to adjust their opinions to a fixed matrix, e.g. a scale. Consider the Eurobarometer surveys (EC 2008, 2009, 2012), which contained the following question: “Would you be comfortable with having a homosexual (gay, lesbian, bisexual) person in the highest elected political position? (scale 1-10, from very uncomfortable to totally comfortable)”. People’s opinions are usually rather vague and ambiguous, while such scales suggest that they are definite, precise and explicit. Associated with this is the problem that small changes in the formulation of the questions may lead to major differences in the answers (Potter and Wetherell 1987: 40). The Eurobarometer question using another scale (for example a range of answers – “very uncomfortable, uncomfortable, rather uncomfortable, rather comfortable, comfortable, totally comfortable” instead of the 1-10 scale) might paint a significantly different picture of the situation. In addition, “self-contradictory answers within survey research often are regarded as threats to the reliability of the study rather than as signs of variation in the use of language. In opposition to this, variation and self-contradictory answers are taken for granted in discursive psychology and such variations are seen as signs of the use of several discourses” (Jørgensen & Phillips 2002: 122). 

 Although this project regards the majority’s attitudes, problems associated with measuring opinions of the minority may also be mentioned. First of all, it is impossible to measure statistically the proportion of the population that identifies as belonging to religious/sexual minorities, due to their stigmatization by the society (non-heterosexual persons are assumed to constitute 5-10% of the population; Glossary of Terms in Makuchowska (2011: 11). Many of their members lead double lives and their minority identities are invisible. Second, if they do identify publicly and agree to participate in a survey, their answers are, of course, subject to all reservations discussed above. The use of, for example, police records for assessing the extent of discrimination or hate crimes is also problematic. Many authors writing about the social situation of sexual minorities emphasize that most of them (even up to 90%) do not report discrimination or physical/psychological violence against them to the police (e.g. Jugovič, Pikič & Bokan 2007; Makuchowska & Pawlęga eds.) 2012).

No comments: