[ Pobierz całość w formacie PDF ]
.But we do not know why Russians have this mixtureof attitudes.The usual explanation that Russians inheritednondemocratic attitudes from their own cultural past, butlong-term social change is now changing some of thoseideas does not cover all the facts.It is not only older and lesseducated Russians who worry about democracy and markets.It is not clear that popular attitudes have been changing overtime, even under the influence of a highly turbulent present.Explanations that divide the Russian population into supportersand denigrators of democracy also do not help.In Russia, it isoften not the case that some people hold consistently democraticattitudes while others are more authoritarian.Instead, manyindividuals seem to hold a mix of apparently contradictory ideas.Opinion surveys can tell us a great deal, but they cannottell us why many Russians hold what seem to be inconsistentideas about political life.For that explanation, it is helpful toallow ordinary Russians to speak in their own words.Inten-sive interviews, in which respondents can speak at length,tell stories, give examples, and generally explain the logicof their ideas, are a valuable tool in sorting through some ofthe conflicting findings and contradictory images that survey3 01.Carnaghan FM-Ch4 1/17/07 12:13 PM Page 4444 Out of Orderresearch in Russia has produced.In this chapter I explain the intensiveinterview method and compare what we can learn from intensive inter-views with what we can learn from standardized opinion surveys.Whilesurveys are essential in drawing the big picture of attitude distributionacross a countrywide population, intensive interviews can help determinewhat that big picture means.Drawbacks of StandardizationStandardized public opinion surveys provide outstanding data for a numberof purposes, including, particularly, generalizations to the wider populationand multivariate data analysis.But at the same time, these data have certainlimitations.Some of the limitations stem from the standardization inherentin a large-sample survey.In order to be able to analyze the answers of athousand of people or more, survey researchers have no choice but totransform their research concerns into multiple-choice questions.Eventhe responses to open-ended questions must be grouped into a relativelysmall number of categories if the goal is to conduct statistical analysis.Respondents cannot talk with the interviewer as if the two were involvedin a normal conversation.Rather, they must respond to predeterminedquestions with a choice of preset answers.In most survey protocols,respondents cannot reframe the questions, elaborate on or explain theirresponses, or even receive substantial clarification about confusing items.So as not to influence results, interviewers must ask the same questions,in the same order, with the same neutral inflection, of all respondents.This standardization makes it possible to summarize the answersof thousands of respondents and to examine statistical relationshipsbetween the various answers.Just as there are benefits, though, there arealso costs of such standardization.Respondents may find the artificialityof the interview experience off-putting.Some researchers worry thatrespondents may psychologically withdraw from the interview processonce they realize that only small pieces of their opinions are of interest tothe interviewer.1 If the artificiality of the interview situation stunts conver-sation, it can also limit the information that researchers get from surveyrespondents.The short answers respondents are allowed to provide may1.Lucy Suchman and Brigitte Jordan, Validity and the Collaborative Construction ofMeaning in Face-to-Face Surveys, in Questions About Questions: Inquiries into the CognitiveBases of Surveys, ed.Judith M.Tanur (New York: Russell Sage Foundation, 1992), 244. 01.Carnaghan FM-Ch4 1/17/07 12:13 PM Page 45Russians in Their Own Words 45be misleading if they leave out the reasoning that led to the answer or thequalifications that substantially moderate the recorded response.Theanswers will also be misleading if the respondent has lost interest in theinterview process and is answering questions by rote.The accuracy of standardized survey responses also depends onrespondents and survey researchers sharing the same interpretations ofwords.But we cannot be certain that respondents understanding of wordsand concepts matches the intention of the researcher.Public opinionresearch has shown that even words such as children, you, weekend, andgenerally do not have a universally shared meaning.2 Certainly the same istrue for government, trust, or extremist, words commonly used in ques-tions about democratic values or responses to political institutions.Thedifficulty of knowing what words mean is compounded for Americanresearchers interviewing Russians.Some words in English do not havedirect equivalents in Russian.Others have similar meanings to those of avariety of Russian words, which in turn do not all mean the same thing toRussians.In recent surveys of Russians by Western scholars, the wordgovernment has been variously translated as pravitel stvo (the cabinetministries), or as gosudarstvo, literally state, as in Moscow State Univer-sity.Similarly, people is variously translated as the neutral liudi or as narod, a word that may designate the common people or the ethnicallyRussian people, but that probably does not include everyone in the country.3These differences most likely influence results and can explain some ofthe variation in findings between surveys.Another problem that stems from the standardization of large-N surveysis that respondents have to be able to select the answer that most closelyreflects their own opinions.That may not be too difficult if the respondentunderstands all the alternative responses and has a precisely articulatedopinion that nearly matches one of the alternatives provided.But it is farfrom obvious that this is normally the case.Inspired by the instability ofsurvey responses over time and their sensitivity to phenomena such asquestion order, some public opinion researchers have concluded thatmuch of the public does not actually have opinions.4 Russian researchershave echoed this concern, on the logic that people who for decades were2.William Belson, Validity in Survey Research (Aldershot, U.K.: Gower, 1986).3.Nancy Ries, Russian Talk: Culture and Conversation During Perestroika (Ithaca: CornellUniversity Press, 1997).4.Philip E.Converse, The Nature of Belief Systems in Mass Publics, in Ideology andDiscontent, ed.David E.Apter (New York: Free Press, 1964), 206 61. 01.Carnaghan FM-Ch4 1/17/07 12:13 PM Page 4646 Out of Ordernot allowed to voice opinions may not ever have developed them.5 Theseconcerns are probably exaggerated, but evidence suggests that most peoplepossess multiple and often conflicting opinions that do not fit easily intothe narrow framework of a standardized questionnaire.6 The answers thatrespondents provide to closed-choice questions seem to depend on whatwords are used, what questions occurred earlier in the interview, and theattributes and mannerisms of the interviewer.Confronting respondentswith counterarguments or other information often induces them tochange their minds.7 Survey responses are thus snapshots of reality, butthe reality they describe may have seemed quite different had the shotbeen taken from a slightly different angle
[ Pobierz całość w formacie PDF ]
