Qualitative Method of Semi-Structured Interviews - Essay Example

Qualitative Method of Semi-Structured Interviews – An Individual Research Introduction Semi-structured or focused interviews are often used in qualitative research, particularly in the field of media research. The questions are contained in an interview guide (not 'an interview schedule' as in quantitative research!) with a focus on the issues or topic areas to be covered and the lines of inquiry to be followed. The sequencing of questions is not the same for every participant as it depends on the process of each interview and the responses of each individual. The interview guide, however, ensures that you collect similar types of data from all informants. In this way, you save time and the 'dross rate' is lower than for unstructured interviews. The interview guide allows you to develop questions prior to interviewing and then decide for yourself which issues to pursue. Example An interview guide The following are some questions, which aim to find out about the process of setting up a new, public relations consultancy: • How did you come to start up the online-PR company? • How did you know how to go about managing a commercial organization? • How did you feel about taking on all this responsibility at such a young age? • What happened when you gained your first client? • How did you know what online methods to employ to promote that client's type of business? And so on … Although the interview guide may be quite long and detailed, it need not be followed strictly. It should, however, focus on particular aspects of the subject area to be examined, but can be revised after interviewing because of the ideas that arise. Although the aim is to gain the perspectives of informants, you need some control over the interview (which the guide provides) so that the research topic can be explored and the purpose of the study achieved. Ultimately, you must decide what interview techniques or types are best for you, the topic and the interview participants. To the extent that observation occurs, it is largely non-participant, with the researcher being very much on the periphery of interaction, and is undertaken in a somewhat unstrategic manner. Such observation is usually supplementary and something that is carried out in the spaces between interviews or at mealtimes. Burgelman’s (1985) research on a new venture division, which relied primarily on unstructured interviews and the examination of documents, is an example of an interview-based study. Bryman et al. (1988) report the results of a study using semi-structured interviews, which sought to examine the perceptions of leadership among individuals in managerial and supervisory positions on three construction projects in the UK. One strand in the investigation was the kinds of factor that is perceived by managers as affecting the styles of leadership that they need in order to enhance project performance. The bulk of research on leadership effectiveness has tended to emphasize the role of broad categories, such as task structure or leader’s position power, as situational factors that must be taken into account in seeking to maximize performance (Bryman, 1986). Bryman et al. found that their respondents emphasized factors that were specific to the construction industry. For example, because construction projects are temporary, time perspective plays an important role in leaders’ decisions about the most effective style. As one respondent put it: I think a fast job, when there is no room for messing about; you get a much more directive style of management. I think if you get time on the job, it’s a more consultative style of management, you are more willing to get blokes in to sort the thing out, be together, you’ve got time for a joint approach. On a fast job, you cannot afford that. (Bryman et al., 1988, p. 22) This study involved very little non-participant observation. When such observation occurs in interview-based studies, its role is mainly to augment interviews and documentary materials (Collinson, 1988; Jackall, 1983) and often assists in the formulation of interview questions (Burgelman, 1983). Gouldner (1954) found that the interviewing of miners was made easier by periods of observation, since it enhanced his ability and that of his research team to talk meaningfully with their respondents about the work. While the studies in the interview-based group tend, as their name implies, to emphasize interviewing, a study by Pettigrew (1985) of organizational change in four divisions and the head office of Imperial Chemicals Industries (ICI) in the UK made considerable use of documents. In this research, Pettigrew looked in particular at the intervention and effects of organization development (OD) consultants. The research was specifically concerned to elucidate processes over time, so an explicitly longitudinal approach was designed. Over the period 1975 to 1983, Pettigrew interviewed 134 ICI employees (some more than once). The interviews were semi-structured and derived from Pettigrew’s reading of the relevant literature and from conversations with some ICI employees to produce ‘a very open and flexible set of themes and questions’ (1985, p. 40). Some 500 hours of tape-recorded interviews were produced. Crucial documents concerning the company were scrutinized (such as documents relating to company strategy, to organizational changes and to the training and use of OD consultants), while informal conversations and some observation in factories and offices provided additional data. The period covered by Pettigrew’s research was 1965 (the year in which OD work got under way) to 1982. Thus, some of Pettigrew’s data derive from retrospective questioning of key figures (covering the period 1965 to 1975), but from 1975 the data are (as he puts it) in ‘real-time’. A broadly similar approach can be discerned in Johnson’s (1987) study of strategic change at Foster Brothers, a menswear retail chain. Through the combination of retrospective and concurrent interviewing over a number of years and past and present high-level documentary evidence, these studies constitute a form of interview-based qualitative study in which an explicit longitudinal emphasis is thoroughly exploited. Multi-site Studies Multi-site Studies entail the collection of qualitative data in a number of different organizations, unlike research subsumed under the other categories, which rarely involves the collection of data in relation to more than two organizations and in any case not in excess of five. The methods are fairly varied; Isenberg (1984) and Kotter (1982) employed a combination of observation and interviews, Sutton (1987) interviews and documents and Bennis and Nanus (1985) and Goss (1988) interviews alone. Because of the researcher’s involvement in a number of organizations, the amount of time spent in each is generally considerably less than that exhibited by the three other types. Isenberg (1984) spent a mode of two and a half days with each of the dozen executives he investigated. Sutton (1987) carried out a maximum of eight interviews in each of the dying organizations he studied. Kotter’s (1982) research entailed a number of different methods of data collection. He observed fifteen general managers (GMs) in a largely unstructured manner for a total of 500 hours. He spent nearly 100 hours interviewing them and 200 hours interviewing people with whom they worked. In addition, relevant written documents were accumulated and two questionnaires were administered, the latter largely to provide data on background and personality. To provide a flavour of the observational evidence collected by Kotter, the following record from a conversation between one GM (Thompson) and two of his subordinates (Dodge and Smith) is provided: THOMPSON: What about Potter? DODGE: He’s OK. SMITH: Don’t forget about Chicago. DODGE: Oh yeah. (Makes a note to himself.) THOMPSON: OK. What about next week? DODGE: We’re set. THOMPSON: Good. By the way how is Ted doing? SMITH: Better. He got back from the hospital on Tuesday. (Kotter, 1982, p. 89) This passage, which has been extracted from a much longer conversation, is utilized by Kotter to illustrate that many of the conversations he witnessed appear highly chaotic, but in fact were extremely efficient. Four tasks were achieved in a matter of seconds: 1 He learned that Mike Potter has agreed to help on a specific problem loan. That loan, if not resolved successfully, could have seriously hurt his plans to increase the division’s business in a certain area. 2 He reminded one of his managers to call someone in Chicago in reference to that loan. 3 He also learned that the plans for next week, in reference to that loan, are all set; these included two internal meetings and a talk with the client. 4 He learned that Ted Jenkins was feeling better after an operation. Ted worked for Thompson and was a reasonably important part of his plans for the direction of the division over the next two years. (Kotter, 1982, p. 90) These passages illustrate well the potential of the kind of data that derives from the use of observation (usually a ‘fly-on-the-wall’ approach) in multi-site studies. When interviewing is conducted, it is usually to provide information about a focal individual (for example, Kotter) or about organizations (for example, Sutton). Further, inferences are made across individuals or organizations; the GM, Thompson, is taken to exemplify a pattern or tendency common to most of the GMs he observed. Although multi-site studies could be viewed as veering in the direction of survey research, the search for common features is consistent with Van Maanen’s (1983, p. 257) suggestion that ‘qualitative work is more concerned with commonality and things shared in the social world than it is with differentiation and things not so shared’. Comparing the four types All social research is subject to trade-offs; even when researchers do their hardest to dovetail research problem and research method or design, it is rarely possible to avoid certain pitfalls or disadvantages in the choice(s) made. In reality, much research entails an attempt to maximize ‘damage limitation’. Similarly, the four types delineated thus far allow the concerns of qualitative research to be realized in different ways and degrees. The seven distinguishing features of qualitative research can serve as a framework. These features are a matter of considerable importance to qualitative researchers, many of whom have seen their tradition as a movement to challenge the orthodoxy represented by quantitative research (Morgan and Smircich, 1980). First, while all four types display a concern to forge interpretations in actors’ own terms, the degree to which this occurs varies considerably. Van Maanen (1982) has observed that while the insider/outsider distinction is often employed to distinguish qualitative and quantitative research, it can also be used within the qualitative tradition. An outsider perspective in qualitative research is likely to occur when ‘the framework selected by the researcher to interpret what is observed is not one that would be thought obvious, natural, or probably very interesting by those studied’ (Van Maanen, 1982, p. 18). The reliance on interviews in interview-based and some multi-site studies might be taken to indicate that the degree of penetration of, and fidelity to, the perspectives and interpretations of those studied is less pronounced (since the researcher’s participation and involvement in the organization are less marked) than in total and semi-participant studies. Except in the case of the most unstructured interviews, some kind of guide or schedule is produced which is bound to reflect the preconceptions of researchers, even though they may try very hard to detect and follow up subsequently those matters that are of particular concern to their interviewees. However, the outsider element can be discerned in other study types. Pinfield’s (1986) total participant study involved an attempt to test and evaluate two views of the decision-making process: the structured perspective that views it as a rational, orderly sequence and the anarchic perspective that views decision-making as a much more random, haphazard activity. However, this way of thinking about decision-making probably departs quite considerably from participants’ perspectives. Therefore, although there may be a tendency for total and semi-participant studies to exhibit greater fidelity to members’ interpretations, there can be exceptions. The contextual emphasis tends to be less pronounced in the multi-site studies than in the rest. The requirement of developing common themes across a number of cases often entails a sacrificing of the contextual nuances that are usually evident in the other types. Third, there does not seem to be anything inherent in any of the types that inhibits the infusion of a processual element, but it tends to be handled in different ways. In total and semi-participant studies, processes, in the sense of interconnections of events over time, tend to be observed as they happen; in interview-based and multi-site studies, processes are usually inferred from documents and retrospective questioning of informants. Pettigrew’s (1985) ICI research is an exception in that aspects of change could be established through ongoing interviewing. An example of retrospective questioning leading to clear inferences about processes is Sutton’s (1987) study in which a model is produced that depicts the interpretation shifts that occur as members of dying organizations initially view their organizations as permanent but struggling, then as dying and then recognize that death has occurred. The role of leaders in bringing about these interpretation shifts is articulated within this process model. Fourth, multi-site studies are probably the most structured of the four types, because of the need to draw reasonably comparable data across different cases. Interview-based studies tend to be somewhat more structured than total and semi-participant studies, whose broadly unstructured approach means that they usually exhibit greater flexibility. On the other hand, in his interview-based study of one of five construction projects, Bresnen (1988) notes how some chance remarks in a pub, which implied some conflict between senior and junior management, which he had not picked up previously, prompted a new line of inquiry. Fifth, virtually all of the studies that have been examined employed more than one data source, so that the use of multiple methods of data collection does not seem to be associated with a particular study type. Sixth, total and semi-participant studies probably display a greater emphasis on the notion that organizational reality is something constructed by individuals rather than a prior entity that is independent of people. Barley’s (1986) study of the introduction of CT scanners emphasizes that the contrasting social structures of the two radiology departments in which he was an indirect participant observer could be attributed to the different patterns of reaction to the new technology as experienced by those personnel affected by it. The social structures were not simply ‘there’; they were an outcome of the way in which the technology was integrated into the contrasting social arrangements of the two departments. However, a multi-site study like Sutton s is able to show how organizational definitions of the state of health of dying organizations are substantially affected by leaders’ attempts to alter members’ perceptions, suggesting that other study types can incorporate this aspect of qualitative research. On the other hand, interview-based and multi-site studies generally seem to place less emphasis on this feature. Finally, total and semi-participant studies generally exhibit much greater proximity to, and first-hand involvement with, the flow of interaction. In interview-based and some multisite studies, the reliance on interviewing removes the researcher somewhat, creating a layer between researcher and subject. Conclusion In summary, studies in which there is a substantial amount of participant observation (like media related research); probably exemplify the general characteristics of qualitative research better. These are precisely the characteristics on which the proponents of qualitative research have sought to emphasize the uniqueness of their enterprise (Morgan and Smircich, 1980; Rosen, 1986b). However, not all topics and contexts lend themselves equally to a total or semi-participant approach: it is difficult to imagine using such an approach in connection with the long time-spans covered by Pettigrew (1985), for example. Further, as far as media studies go, interview-based and multi-site studies carry their own special advantages, such as the greater opportunity for studying a number of organizations (and hence potentially greater generalizability) offered by multi-site research. Bibliography Barley, S.R. (1986), ‘Technology as an occasion for structuring: evidence from observations of CT scanners and the social order of radiology departments’, Administrative Science Quarterly, vol. 31, no.1, pp. 78-108. 155, 160, 171 Burgelman, R.A. (1985), ‘Managing the new venture division: research findings and implications for strategic management’, Strategic Management Journal, vol. 6, no. 1, pp. 39-54. 25-6, 29-30, 135, 141, 152, 155, 172 Bryman, A., Bresnen, M., Beardworth, A., and Keil, T. (1988), ‘Qualitative research and the study of leadership’, Human Relations, vol. 41, no. 1, pp. 13-30. 152, 155-6 Bryman, A. (1986), Leadership and Organizations (London: Routledge & Kegan Paul). 65, 74, 124, 133, 156, 190, 231, 245 Burgelman, R.A. (1983), ‘A process model of internal corporate venturing in the diversified major firm’, Administrative Science Quarterly, vol. 28, no. 2, pp. 223-44. 141, 152, 156, 172-3 Bryman, A., Bresnen, M., Beardworth, A., and Keil, T. (1988), ‘Qualitative research and the study of leadership’, Human Relations, vol. 41, no. 1, pp. 13-30. 152, 155-6 Bennis, W.G., and Nanus, B. (1985), Leaders: The Strategies for Taking Charge (New York: Harper & Row). 152, 157 Collinson, D.L. (1988), ‘ “Engineering humour”: masculinity, joking and conflict in shop-floor relations’, Organization Studies, vol. 9, no. 2, pp. 181-99. 152, 156 Gouldner, A.W. (1954), Patterns of Industrial Bureaucracy (New York: Free Press). 147-8, 152, 156, 170-1 Goss, D. (1988), ‘Diversity, complexity and technological change: an empirical study of general printing’, Sociology, vol. 22, no. 3, pp. 417-31. 152, 157 Isenberg, D.J. (1984), ‘How senior managers think’, Harvard Business Review, vol. 62, no. 6, pp. 81-90. 152, 157 Jackall, R. (1983), ‘Moral mazes: bureaucracy and managerial work’, Harvard Business Review, vol. 61, no. 5, pp. 118-30. 152, 156, 161 Kotter, J.P. (1982), The General Managers (New York: Free Press). 152, 157-8, 227 Morgan, G., and Smircich, L. (1980), ‘The case for qualitative research’, Academy of Management Review, vol. 5, no. 4, pp. 491-500. 27, 159, 161, 169, 248, 250, 253 Pinfield, L.T. (1986), ‘A field evaluation of perspectives on organizational decision making’, Administrative Science Quarterly, vol. 31, no. 3, pp. 365-88. 152-3, 159, 166, 174, 253 Pettigrew, A. (1985), The Awakening Giant: Continuity and Change in ICI (Oxford: Blackwell). 152, 156-7, 160-1, 197, 243 Pettigrew, A. (1985), The Awakening Giant: Continuity and Change in ICI (Oxford: Blackwell). 152, 156-7, 160-1, 197, 243 Rosen, M. (1986), ‘Some notes from the field: on ethnography and organizational science’, Dragon, no. 6, pp. 57-77. 146-7, 151-4, 161, 163 Sutton, R.I., and Callahan, A.L. (1987), ‘The stigma of bankruptcy: spoiled organizational image and its management’, Academy of Management Journal, vol. 30, no. 3, pp. 405-36. 171, 173, 178 Semi-structured and Non-Directed Interviewing Techniques'. Marketing Intelligence and Planning, 14(6), 6-15. Read More