Data collection through open semistructured interviews

Systematic interviewing is a principal data collection technique in qualitative and applied research (Mantwill et al. 1995:68-9, Robson 1993:228-31, Ackroyd and Hughes 1992:100-4, Strauss and Corbin 1990:18, Patton 1990: 10, Yin 1989:19, Sampson 1986:32-3, Merton and Kendall 1945-6). Utilizing the language of the consumer is an essential requirement of empirical research that investigates a cognitive research problem such as that raised by this study: semi-structured, open interviews were conducted with individual consumers.

Research into intended behaviour versus actual behaviour

Consumer cognition can be approached through interviews that enquire either about intended, future behaviour or about actual, past behaviour:4 the latter approach, which is favoured by this study, draws on retrospective recall, which means that interviewees have to search their memories for recollections, while the former approach draws on hypothetical reasoning.

Research into intended, future behaviour may be biased because of its hypothetical and/or prospective character. When asked, consumers might readily say that they would shop in an environmentally friendly way in future. The cognitive complexities and practical difficulties of 'going green' and overcoming them might be underestimated or wrongly judged because of the lack of actual experience. A gap between intended and actual behaviour could result, and an interview technique based on hypothetical reasoning has not much to offer to deal with such a gap.

Intention research into green behaviour may also elicit responses from consumers on the basis of a heightened general environmental awareness only, e.g. a heightened concern about topical environmental problems such as global warming. Such a general awareness can normally not be expected to lead (directly, imminently or unavoidably) to specific behaviours, such as green consumption or green investment. An observed 'words—deeds inconsistency' between intended behaviour and actual behaviour can be partly attributed to misguided attempts to deduce specific behaviours from a general awareness (see chapter 1). The Fishbein model clearly states that only attitudes and intentions regarding a specific issue but not a general one determine actual behaviour (Ajzen and Fishbein 1980, Fishbein and Ajzen 1976, Fishbein and Ajzen 1975; see also chapter 1). In addition, an intention-behaviour gap may partly be due to a social desirability bias. Voicing non-green buying intentions during an interview might be experienced as socially embarrassing (Mintel 1992:2, MacKenzie 1991:71, Wimmer 1988:76).

Actual behaviour appears to play an important role when it comes to knowledge development and the build-up of experience: 'Knowing is doing' (Maturana and Varela 1992:244-8; see also LaPiere 1934:2367). Research on past behaviour that draws on retrospective recall has been successfully applied to consumer behaviour research in general, and to cognitive research of repeat purchase decisions in particular. This technique can be used when prior knowledge—experience—exists (MacInnis et al. 1992, Zeithaml 1991, Eisenhardt 1989a, Brucks 1986, Hoyer 1986, Smith and Houston 1986, Reynolds and Jamieson 1985, Duncan and Olshavsky 1982, Bettman 1979b, Markus 1977). Major market researchers, such as MORI or Mintel, have also researched green consumer behaviour with regard to past behaviours rather than intended ones (Worcester 1995:7-9, Upsall and Worcester 1995:8-10, Worcester 1993a: 318, Worcester 1993b, MORI 1992, Jacobs and Worcester 1991, Mintel 1991).

Interview biases

The kind of biases from which intention research tends to suffer present less of a problem for research of actual behaviour through retrospective recall. A general advantage of a research procedure that draws on retrospective recall lies in the fact that it does not interfere with the cognition and behaviour that occurred in its real-life context in the past. However, other biases have to be checked for. Retrospective recall may be affected by bias because of inferences in recall—a so-called 'theorizing' or 'rationalizing' bias—and a cognitive retrieval bias related to the phenomenon of forgetting.

Ideally, verbal statements of interviewees should reflect past cognition and behaviour. Any question asked by an interviewer presents an interviewee with information that is meant to stimulate a verbal response. Inferences in recall refer to the change in knowledge structures as 'new reasoning patterns' are applied to 'old behaviours' at the point of questioning, e.g. a change in the abstractness of cognitive processing might occur (Johnson and Fornell 1987: 219, Brucks 1986:60, Bettman 1979b:196). The creation of new knowledge that is due to inferences in recall is undesirable. Through the structuring of an interview, in particular through the way in which retrieval of knowledge is organized and questions are sequenced, an inferences in recall bias can be minimized. One advantage of open interviews over open mail questionnaires is that the sequencing of questions is likely to be more effective. In the case of an open mail questionnaire, people can skim through before answering it, and thus sequencing effects that might help to restrain inferences in recall are eliminated. This issue will be returned to when the interview procedure is outlined.

Some cognitive research projects have deliberately structured interviews to force 'the consumer up the ladder of abstraction' (Zeithaml 1991:28, also Gutman and Alden 1985:103-4, Reynolds and Jamieson 1985:121-2) in order to gain an understanding of his or her cognitive operations. Such an approach deliberately attempts to utilize inferences in recall. Its highly manipulative and hypothetical nature makes it inappropriate for capturing the actual standpoint of the green consumer, which is attempted in this study. However, a similarly hypothetical approach was applied by this study to a comparison group of so-called 'non-green' consumers (subsequent sections give more details on how non-green consumers were researched).

Besides inferences in recall, the phenomenon of forgetting can present a problem for any empirical research technique that enquires about past cognition and behaviour (Lynch and Srull 1991:104). With regard to the product focus chosen by the study, this problem may be comparatively trivial. It has been found that recall is enhanced by the recency of behaviour, by its frequency, and by the involvement of the customer (Engel et al. 1990:509, Alba and Hutchinson 1987:434). For the buying of everyday products, such as groceries, recency and frequency are common characteristics. High involvement is normally not associated with the buying of products for daily consumption, but for green shopping some sort of involvement is likely to exist.5

The problem of forgetting may be reduced by prompting consumers, i.e. by asking highly specific questions that draw on so-called part listings, e.g. lists of certain products. A disadvantage of part listings is that they restrain recall and lead to a prompting bias since they focus recall narrowly. Through part listings, the nature of a recall task is changed from a free recall task, where there is no prompting, to a recognition task (Lynch and Srull 1991:107, 113-14, Alba et al. 1991:6, 19, Nickerson 1984:531-2). Hence, the positive effects of part listings that help to reduce the problem of forgetting have to be balanced against their negative effects in restraining and focusing recall. Such a trade-off has to be kept in mind when the interview structure is designed.

0 0

Post a comment