O socjologicznym wywiadzie kwestionariuszowym: psychologia poznawcza Norberta Schwarza i jej badawcze zastosowanie w warunkach polskich

Abstract

The questionnaire interview with which we deal in this book is a data collection technique characterized by a high degree of standardization. This characteristic is seen as an advantage by some and as a fault by others, often provoking cautious, if not condemnatory judgements, since the questionnaire interview is associated with an „unnatural”, artificial, or even hypocritical situation, where interviewers ask for opinions on matters often unfamiliar or not understood by respondents and they in turn provide answers that do not reflect their knowledge and the degree of opinion crystallization. We personally adopt an approach of a methodologist-realist who, taking into account the actual situation of his own discipline and specific conditions for conducting empirical research pertaining to this discipline, notices that although the questionnaire interview is burdened with many faults, it still has a number of advantages. Recognizing the questionnaire interview as a legitimate method in the “arsenal” of sociological 'research techniques, a methodologist-realist not only assesses its strengths and weaknesses realistically, but also checks its effect in special experiments – trials and tests. It should be strongly emphasised that realistic methodology of the questionnaire interview does not have any paradigmatic affiliation – it does not belong to any trends of sociological thought or sociological research programs. Neither is it unavoidably the “error methodology”, as it was a little maliciously called long ago. Having adopted the positivistic epistemological approach, the “error methodology” uses the concept of measurement based on the assumption of real value and a tool exposing itself to errors in establishing it, in particular referring to the idea of questions in an interview questionnaire leading to answers that are not always true and therefore require verification. In its epistemologically weaker form – and this version is discussed further in our book – it may be the “effect methodology” and as such it talks, with more restraint, only about differences and not measurement errors in the cognizance of reality – and obtained pictures of reality caused by the use of different research techniques, e.g. different forms of questions in a questionnaire or their layout “Error methodologists” are especially interested in these differences, which they relatively define as “effects”. Such “effects” can be observed, for example, when answers of a certain type are more prevalent among answers to a question formulated in form A than among answers to a question formulated in form B. The discovery that form A, as opposed to form B, produces a specific effect does not necessarily lead to the conclusion that it is better or worse, burdened with a smaller or bigger error, it is just enough to say that it is different and causes a characteristic “overcolouring” or a peculiar “leaning” in the picture of reality. The “effect methodology” satisfies itself with a test which does not intend to establish the truth and thus does not verify, but only checks what happens to responses and in what circumstances and how it affects research results. When it comes to the analysis of a questionnaire interview, and especially the relations between a question and a response, the similarity of both these methodologies can be seen in their characteristic tendency to not only describe “errors” or “effects”, but also to interpret and explain them, especially by looking into a respondent’s psychology. Both for Jan Lutyński’s methodology, based on the concept of “error"’, and methodology of Norbert Schwarz, who uses the concept of “effect”, it is important to decipher the “black box“ which is the mind of a person answering the question. The hope for such explanatory success in both these methodologies is justified and maintained in a slightly different way. Schwarz, whose work we discuss in this book, bases his hope on the belief in the theoretical power and empirical usability of cognitive psychology. He expects that cognitive psychology will allow to understand the mental processes taking place in a respondent’s mind when he receives questions and answers, them, revealing – on individual level – the mechanisms “responsible” for the obtained effects. What is more, Schwarz hoped that such application of cognitive psychology in survey methodology would make it possible to transform “the art of asking questions” into a "theory of asking questions”. Cognitive psychology examines how humans acquire, process, store and use their knowledge. It assumes that a man’s behaviour and his motivational and emotional processes depend on the information received by his mind. To understand or predict a man’s behaviour, first one has to understand how he perceives phenomena, what emotions he feels towards them and what actions he undertakes in connection with them. Pieces of information about these three aspects of reality, i.e. fragments of the world are recorded in memory and are called cognitive schemata, cognitive structures or cognitive representations. A man uses these schemata to compare incoming information with information already contained in them, taking decisions about its congruence. New data are encoded in memory in the same place and with the same label as information similar to them. Constantly changed and updated, schemata are used as a basis for opinions, decisions, conclusions and predictions. In his work, Norbert Schwarz combines methodology of the questionnaire interview with “applied” cognitive psychology. On his own and with the help of his associates (George Bishop, Herbert Bless, Hans-Jürgen Hippler, Thomas Münkel, Elisabeth Noelle-Neumann, Fritz Strack, Seymour Sudman, Michaela Wänke), he carried out innumerable experiments, demonstrating the presence of different effects in the results of studies using a questionnaire interview technique (or other techniques based on communication). The experiments were designed to detect the effects of: primacy, recency, contrast, context (assimilation or contrast) and scale. Moreover Schwarz explored the following issues: the forbid-allow asymmetry, comparative judgments, offering the “don’t know” response, changing the position of the middle alternative. In our book we address the following question: if Schwarz’s experiments were repeated here in Poland, would they yield the same results? What would change in his findings if, in this comparison, the form of questions or their sequence remained the same but their content changed fundamentally, adjusting it – which is probably quite natural – to Polish conditions? At the basis of so formulated question lies the conviction shared with cognitive psychologists that people process their knowledge about the world in the same way, i.e. by using cognitive schemata, regardless of their individual qualities or characteristics of the place and times in which they live. At the same time, these schemata have a universal character and so are not influenced by e.g. age, gender, place of residence or nationality. Comparing the results of studies from Germany and the United States Norbert Schwarz does not wonder if society in which subjects live and its culture have an impact on their answers. Just as mechanically, or even unthinkingly, he compares the results of studies conducted on purposefully selected groups of students and respondents from representative random samples. Having accepted the logic of his studies we still want to know if experimental manipulations similar to those carried out by Schwarz will produce the same effects here as the ones obtained by him. To verify this hypothesis in September 2001 we conducted our own study on a randomly selected sample of adult inhabitants of Łódź using a questionnaire interview. The study was planned as an entire series of methodological experiments. The questionnaire was prepared in two versions – A and B. Both versions were identical in terms of information that we were looking for. Each consisted of six thematic sections (concerning: attitude to law, assessment of safety, equality of citizens to the law, opinions on abortion, homosexuality, divorce; alcoholism, drug addiction, attitude to the sick, political issues, free time, sense of happiness) and background questions. They differed, on the other hand, in the construction of individual questions, the order of questions in a section and the order of particular sections. Respondents in our study were randomized into two subsamples (corresponding to two versions of the research tool): persons with odd numbers on the list were administered version A of the questionnaire, whereas subjects with even numbers were asked questions from version B. Then we screened the completed questionnaires, choosing 451 for further analysis (229 for version A and 222 for version B). Detailed analyses revealed that these subsamples could be treated as equivalent, i.e. respondents’ answers to the questions in both versions of the questionnaire were comparable. In the table below we present the number of conducted experiments including those confirming and those failing to confirm the effects obtained by Schwarz. As it can be seen most experiments adapted to Polish conditions did not confirm Schwarz’s findings. [SEE summary_A_Krzewińska.pdf] The book consists of an introduction, four chapters and a conclusion. In chapter one we present general information about the questionnaire interview, its basic characteristics and areas of application, examining why it is such a controversial data collection technique – strongly criticised by some and fiercely defended by others and nonetheless still present in sociological empirical studies or even popular. Taking into consideration the alliance of survey methodology with cognitive psychology we also look into three other psychological concepts of man as well as “models of the respondent” characteristic for their research applications. At the start of chapter two, we discuss the beginnings of cooperation between cognitive psychologists and survey methodologists. Then we outline cognitive concepts in psychology concentrating only on those issues which are needed to understand the approach adopted by Norbert Schwarz in the analysis of the questionnaire interview. Next we successively deal with cognitive schemata and their role in the process of answering questions in a questionnaire interview, autobiographical memory and its significance for the recollection of past events by respondents during an interview and finally with the models of answering a question in a questionnaire interview, originating from cognitive psychology applications. In chapter three we portray Norbert Schwarz as a researcher who combines cognitive psychology with survey methodology, presenting the main research areas of his interest and his most important scientific achievements. Our main objective, however, is to reconstruct the scientific basis and the style of his experiments, conducted in compliance with the “effect methodology”. Therefore we discuss particular effects which Schwarz is untiringly trying to “track down” in his empirical work, in particular concentrating on the effect of primacy and recency and the effect of contrast and context. We also take into account other issues that he is interested in, such as the significance of the direction of comparisons, the influence of scales on answers given by respondents, the role of the forbid-allow asymmetry, the sense of introducing filter-questions, the impact of emotions on respondents’ answers, logic of conversation or confidentiality in an interview. The results of our methodological study and their comparison with the results obtained by Norbert Schwarz are presented in chapter four. The order of issues presented in this chapter corresponds with the order in which they were described in chapter three. We are trying not only to describe but also to explain the discrepancies between the results of our study conducted in Polish conditions and the results obtained by Schwarz. In these explanations we often have to go beyond the area of Schwarz’s cognitive theories and hypotheses and refer to a concept or observation that also takes into account the essential influence of social and cultural factors.