Interviewing Marijuana Users: Issues and Strategies for a Reflexive Grounded Theory Approach
Pogovori z uporabniki marihuane: problemi in strategije za pristop refleksivne utemeljitvene teorije
Curtis Fogel Geraint Osborne1
Abstract
This paper examines several limitations of Positivist and Constructivist Grounded Theory approaches to interviewing based on reflections from a study of marijuana use by white-collar professionals and graduate students. The limitations discussed include the following: a) finding and categorizing participants based on the researcher's own conception of who the participants are, b) treating the interview respondents as vessels of facts and knowledge that can be elicited through various interview strategies, c) ignoring the mediating role of the researcher and his/her inscription devices in the construction of interview data, and d) treating the data collected in the interview as reports that mirror lived experiences and realities. This paper concludes with a discussion of some possible strategies to overcome these issues.
Keywords: Grounded Theory; Constructivism; Interviewing; Qualitative Research.
1 Dr. Curtis A. Fogel is a Professor of Interdisciplinary Studies at Lakehead
University- Orillia Campus in Ontario, Canada. Dr. Geraint Osborne is a Professor of Sociology at the University of Alberta-Augustana Campus in Alberta, Canada.
Povzetek
Besedilo obravnava številne omejitve pozitivistične in konstruktivistične pristope utemeljitvene teorije na podlagi opravljenih intervjujev, ki temeljijo na refleksiji študij uporabe marihuane med t.i. 'belimi ovratniki' in diplomiranimi študenti. Obravnavane omejitve vključujejo a)iskanje in kategoriziranje participantov, ki izhajata iz raziskovalčeve lastne koncepcije, b)obravnava respondentov kot gola dejstva in znanja, ki jih izzove pogovor, c)zanemarjanje uravnavajoče vloge raziskovalca ter njegove/njene vloge zapisovanja podatkov v konstrukciji pogovora, d)obravnava zbranih podatkov kot poročila, ki odsevajo življenjske izkušnje in realnosti. Besedilo se zaključi s diskusijo o možnih strategijah, ki bi presegle te omejitve.
Ključne besede: utemeljitvena teorija, konstruktivizem, intervjuji: kvalitativno raziskovanje
Introduction
With much of the emphasis of Grounded Theory approaches placed on coding procedures, theoretical saturation, and theorization, little reflexive attention appears to have been placed on the construction of interview data. The aim of this paper is to explore this construction further by outlining some potential issues with traditional Positivist and Constructivist Grounded Theory approaches to interviewing, as well as to provide some possible strategies for a more reflexive approach to Grounded Theory interviewing. Examples used in this paper are drawn from a study on the use of marijuana by graduate students and white-collar professionals, which used a Grounded Theory approach (Osborne & Fogel, 2007, 2008).
The Grounded Theory approach has been the subject of various
criticisms since its formal inception with Glaser and Strauss's (1967) Discovery of Grounded Theory. Much of this criticism has, however, related to the definition and use of various terms from Glaser and Strauss's (1967) original work, which has now become known as Positivist Grounded Theory because of its attempts to mirror the empirical research methods of the natural sciences. In later works, Glaser and Strauss have made attempts to clarify and further develop their original Grounded Theory text, to reconcile such criticisms (See Glaser, 1978, 1992; Strauss, 1987; Strauss and Corbin, 1998). For more on the general criticisms of traditional grounded theory see the work of Bryman (1988), Silverman (1993), and Bryant (2003).
These criticisms have related primarily to ways of using data collected via traditional Grounded Theory methods. Minimal critical attention appears to have been placed on how data is collected in the first place; instead emphasis has been placed on how the data should be used. This paper makes a significant contribution to this literature by critically examining the ways in which data is initially constructed and developed using a Grounded Theory approach.
In recent years, Constructivist Grounded Theorists, particularly Kathy Charmaz (2000a, 2000b) have provided more detail on conducting Grounded Theory interviews that move away from Positivistic Grounded Theory approaches that ignore the central part of the researcher in the construction of data. However, the mediating role of the researcher, and his or her "inscription devices" (p. 18), still appear understated (Law, 2004). Inscription devices refer to systems or devices that researchers use to produce or trace out materials that take on other forms. Examples of inscription devices could include a pencil to jot down notes, an audio recorder to record, or a computer to transcribe audio to text.
Constructivist Grounded Theorists have made an important turn from a traditional or Positivist Grounded Theory approach by acknowledging that categories, concepts, and theorizations do not merely emerge from data but rather, are defined by the researcher. However, the role of the researcher in the construction of interview data still appears minimal. Furthermore, interview data is still perceived and used as a report that mirrors the experiences and realities of the respondent. This paper encourages a move towards a more reflexive interviewing approach.
Limitations of Grounded Theory Interviewing
The basis of this article stems from methodological issues encountered during the interview process of a study of white-collar professionals and graduate students who use marijuana for non-medical purposes. As the collection of interview data progressed and concluded, a number of possible limitations to Grounded Theory interviewing were apparent. These limitations include the following: a) finding and categorizing participants based on the researcher's own conception of who the participants are, b) treating the interview respondent as a vessel of facts and knowledge that can be elicited through various interview strategies, c) ignoring the mediating role of inscription devices in the construction of interview data, and d) treating the data collected in the interview as reports that mirror lived experiences and realities.
Categorizing Participants
Bruno Latour (2005) a leading figure in Actor-Network Theory, criticizes traditional sociological approaches for privileging the researcher. He argues that this privilege grants researchers the ability to tell actors who they are, what groups they belong to, and how they organize themselves within their groups. This critique is particularly relevant to how interview
respondents are located and categorized within a Grounded Theory approach.
In both Positivist and Constructivist Grounded Theory approaches, interview respondents are typically sought out based on "theoretical sampling" (p.16) procedures (Glaser and Strauss, 1967). A theoretical sample is one that emerges during a study as categories, concepts and theorizations develop; it directs the researcher to interview particular people based on how the study is developing. It is essentially a form of "purposive" (p. 36) sampling, which involves the deliberate selection of subjects who are representative of a given population (Berg, 2004).
Theoretical sampling is considered to be a useful strategy in remaining flexible and open to new possibilities of exploration within a study. Taking Latour's objection into consideration, theoretical sampling does, however, risk privileging the researcher with the capacity to determine who is involved in a particular group. It operates on the presumption that the particular respondent will provide the report that the Grounded Theorist is seeking. The researcher might, however, misidentify the participant, which appears to be a definite challenge with theoretical sampling.
In a study on the subcultural identity of punks, Sue Widdicombe (1998) reveals why it is difficult to label and categorize research participants into given groups. In her study, most of the participants rejected the categorical label of "punk." While Widdicombe (1998) began her work with an idea of who was and was not a member of the subculture of punk, after conducting her research, she found that the respondents did not necessarily confirm these notions. This study reveals the difficulty researchers face in labelling respondents on their behalf.
Group categorization was not a primary issue throughout the study of marijuana use, but certain challenges did exist. The first task was to find people that were marijuana users. This was, however, a difficult category to construct as a large portion of Canadians have used marijuana, although they might not identify as marijuana users. Similarly, we wanted to find students and professionals, which again are not as simplistic of categories to construct as they might appear. Difficulties surrounded the discernment of what work would be considered professional, as well as the categorization of students who were on leave from their professional working position to upgrade their degree. Labelling and categorizing these individuals on their behalf could have negative outcomes on the construction of interview data, as the sample might not reflect how individuals would categorize themselves. By misidentifying participants, it seems problematic that the Grounded Theorist might not be researching the participants that they think they are researching.
Respondents as Vessels of Knowledge
A second limitation of Grounded Theory interviewing encountered in our study was the treatment of the interview respondent as vessels of facts with knowledge that can be elicited through various strategies. While it is certainly important to use a variety of strategies to elicit information from a respondent that might not be discussed otherwise, it appears problematic to ignore the mediating role of the researcher in how these facts, knowledge and experiences are expressed and analyzed.
Grounded Theory was developed by Glaser and Strauss (1967) at a time when American Sociology as a whole was attempting to gain the legitimacy of the natural sciences. The attempt was to create a method that could be considered an objective science, by grounding its method of theorizing within the empirical world. The social world, from this
traditional, positivist perspective, is considered to be what John Law (2004) terms an "out-there reality" (p. 14). An out-there reality exists in isolation to the researcher, waiting to be probed from an objective distance. It is singular and definite.
This description of out-thereness is characteristic of the natural sciences. The biologist, chemist or geologist is said to study a world that exists beyond the researcher. The process of constructing scientific fact rests crucially on the invisibility of the researcher. The goal of a Positivist Grounded Theory approach has been the same; to treat the researcher as invisible in the construction of the social groups and realities to which they are studying.
Bruno Latour (1987, 1988, 1996, 1999), has written extensively on how scientists construct scientific fact. For Latour, scientific fact cannot be constructed independent of scientists and the inscription devices that they employ. From this perspective, it appears that Grounded Theorists have been chasing a false ideal. As Latour would claim, researchers, whether they are chemists, biologists or sociologists, play a mediating role in how data and facts are constructed. For Latour (2005), to ignore this role is to slip further away from objectivity.
Constructivist Grounded Theorists are beginning to acknowledge this mediating role of the researcher in how categories and concepts are constructed. However, minimal attempts have been made to acknowledge the mediating role of the researcher in the initial construction and development of the interview data; instead, the reflexive attention appears limited to how the interview data are used.
This challenge of doing Grounded Theory interviewing was considered at the outset of the marijuana study before any interviews with users were
conducted. To avoid this problem, the interview process was treated as what Mason (2002) describes as a "site of knowledge construction" (p.231). The marijuana users were not perceived as vessels of facts and experiences to be probed, but rather as co-producers of knowledge surrounding the use of marijuana. This allowed for some flexibility to explore issues and concerns prompted by a basic set of open-ended interview questions that were, as should be noted, constructed by the researchers. Just as labelling and categorizing marijuana users on their behalf could have been problematic for the construction of interview data, so too did it appear problematic to feign invisibility in the construction of this data.
Ignoring the Mediating Role of Actants
Acknowledging the part of the researcher in the construction of interview data has become more commonplace after the reflexive turn in sociology (Woolgar, 1988). What has been largely ignored, however, is the mediating role that various technologies or inscription devices play in the construction of interview data. Bruno Latour (2007) suggests that the central aim of Actor-Network Theory is to explore the mediating role that non-humans play within social organizations or networks. He (2005) refers to these non-human actors as "actants" (p. 54). For Latour, nonhuman entities do not become actants simply because they exist, they do so because they "do things" and "make a difference" (p. 154).
Many non-human actors, or actants, appear to be involved in the construction of interview data. For example, Raymond Lee (2004) has explored how recording technologies have shifted the manner to which we conduct interviews. In doing so, the interview data we collect has also changed. Similarly, the process of transcribing audio into text, results in the development of a new actant as the transcribed text may act differently than the audio recording as it may be void of sighs,
pauses, laughter, or tones of exaggeration.
On a more specific level, there are numerous actants that might play a mediating role in the construction of interview data. As illustration, Marjorie Devault (1991, 1999) suggests that women draw on various discourses available to them in how they describe their work of provisioning food for their families. In this sense, the available discourses of both the respondent and the researcher could be conceived of as actants that play an important role in how the interview data is constructed.
Another example, more specific to the Grounded Theory approach, is the actants at play in the process of coding. This process requires verbal speech to be transformed into written text, regardless of what transcribing and coding processes are used. A good example of how transcribed text can be seen as an actant is a close look at how Kathy Charmaz (2000a) describes her technique of coding. Charmaz (2000a), a Constructivist Grounded Theorist, uses what she terms "line-by-line" (p.684) coding of her interview transcripts, whereby she literally divides transcripts line-by-line. It seems apparent that by separating the text by line, different meanings will emerge than if the text were separated by sentence, paragraph, or idea. Essentially, the text comes to act differently based on how it is organized through the coding process.
Like the examples described, the marijuana study had a number of actants that likely shaped how the interview data was constructed. The most obvious actant was the computer and email technology that was used to conduct eleven of the interviews through email. This technology influenced the data that was gathered from these interviews. In some cases it was much more difficult to illicit detailed interaction with respondents through email as compared to face-to-face interviews,
while in others the respondents seemed more comfortable interacting through such a medium. Since actants such as this appear to play an important mediating role in the construction of interview data, it seems to be a disservice to ignore this role.
Capturing Lived Experiences
A fourth limitation of Grounded Theory interviewing encountered in our study was the treatment of the data collected as reports that serve to mirror lived experiences and realities. Part of this presumption is based on the previous three limitations, as the researcher might begin with the assumption that a given participant belongs to a particular group. The researcher might then ignore their own mediating role, and the role of the actants they employ in the interview process, constructing the illusion that the data is formed in isolation of external forces, much like an experiment in a scientific lab. As already shown, such presumptions can be damaging to the use of Grounded Theory interviews.
A further assumption that appears to be held in many Positivist and Constructivist Grounded Theory approaches is that the interview data or transcribed text is able to reflect the lived experiences of the participants. Relatively recent theorizations in Institutional Ethnography have can be used to expose this limitation. A central aspect of many studies using Institutional Ethnography approaches is the exploration of the disjuncture between institutional texts with the lived experiences of people (Smith, 2005). For example, Ellen Pence (1996) explores the disjuncture between women's lived experiences of domestic assault with their institutional representations, whereby much of the women's experiences are left out of the initial police reports and are further tapered as each case travels through the legal system.
The interviewing process of Grounded Theory could be construed as
having a similar tapering effect on the lived experiences of participants. From the outset, the lived experience of the participant is assumed to be accurate and has been mediated by the researcher as well as the various technologies or inscription devices that he or she employs. This lived experience is then textualized in a manner where the researcher determines what is, and is not important, to a given study through processes of transcription, categorizing, coding and theorizing. As the interview data, as a text, travels through these stages, less of the respondents' experiences are captured. The complexity and messiness of the world is reduced into what grounded theorists often refer to as "themes" (Karp 1994, p. 10), or what Latour (2005) cleverly terms "neat little pots" (p. 141).
This limitation provided the most difficult methodological struggle in the study of marijuana use. As outlined, a variety of steps were taken to acknowledge various mediators in the interview process. Through this, we moved the construction of interview data from being "out-there" to "in-here" in John Law's (2004) terms. The interview data was not already in existence, it had to be constructed and developed through a process that involves more than and interviewer probing a passive vessel of knowledge. However, through this process a disjuncture appears to form between the lived experiences of the respondents, and the textual representations of these experiences in the interview data and later theorizings of the Grounded Theorist. A disjuncture still remains between words and worlds, or what Dorothy Smith (2005) refers to as "actualities" and "virtual realities" (p. 2).
Strategies for Reflexive Grounded Theory Interviewing
Despite the various possible challenges of doing Grounded Theory interviewing, various strategies could be used to address and possibly
reconcile these limitations. It is important to note that the term strategies is used here to denote open and flexible means towards more reflexive interviewing, rather than rigid rules that, if followed, will lead to reflexive research. These strategies correspond with the limitations just described. Each discussion will include brief explanations of how they were reconciled in the marijuana use study.
The first challenge of Grounded theory interviewing, to label and categorize interview respondents on their behalf, could be addressed by providing participants with more voice in categorizing themselves. With each step of the research, the Grounded Theorist will inevitably play a mediating role in how the data and analysis come to be constructed, but it might be useful to place more responsibility on the part of the participant throughout these stages. As revealed in the section on limitations, categories can never be taken as a given; even a category as simple as 'student' might not be clear if that participant is a working professional on a leave of absence to take a few courses. Given this, it seems appropriate to have participants categorize themselves as much as possible within a given study.
In the marijuana study, we stumbled upon the use of self-categorization as we were not permitted, via our ethics agreement, to actively seek out research participants given the criminal nature of the topic of study. Instead, we could only have participants contact us if they were interested in participating. Through their expression of interest in the study, they would identify themselves as either student or professional marijuana users. Theoretical sampling was still used until saturation developed, but it was done in a way where participants identified themselves. However, this passive method of theoretical sampling might not be the ideal, as the response rate is often lower when researchers are not able to actively seek out participants (Berg, 2004).
In studies where there are no ethical limitations to contacting participants, it still might be useful to have them label and categorize themselves. For Latour (2004), a central task of sociology is to discern how individuals organize themselves into groups, and not to privilege our selves to do it for them. By having participants categorize themselves, delineating the boundaries of their groupings, a more accurate picture of the subject of inquiry can be developed. By doing so, problems of misidentification on the part of the researcher can be lessened, and a more thorough understanding of how a given group of inquiry is organized and coordinated can be elicited.
The second limitation, treating respondents as vessels of facts, knowledge and experience to be probed at from a distance by an objective grounded theorist, might involve the turning away from scientific ideals. Bruno Latour (2005) has asserted that scientists merely cloak themselves with invisibility despite their clear mediating role in the construction of scientific fact. A strategy towards increased objectivity in Grounded Theory interviewing might then involve increased reflexive acknowledgement of the mediating role of the researcher in the construction of interview data.
The task then becomes to perceive grounded interviewing as what Mason (2002) terms a "site of knowledge construction" (p. 231), rather than as an attempt on the part of the researcher to employ a number of strategies to probe the passive vessel of knowledge that is the research participant. This interviewing approach was adopted in the interviews with marijuana users to limit the false sense of objectivity that characterizes Positivistic, and to some extent, Constructivist Grounded Theory approaches.
An additional strategy to allow for the co-production of knowledge within an interview might be to discuss the formulation of possible interview questions pertinent to the given study with the interview participants, which we experimented with in our study. The tendency of Grounded Theorists, and social science researchers in general, is to formulate interview questions on one's own, then posing these questions to interview respondents. This process can be rigid or flexible in its approach, but the researcher invariably formulates and asks the questions. As a collaborative site of knowledge production, it might be useful to discuss the formulation of other possible interview questions with participants. Such a strategy might elicit interesting interview questions, and would also contribute to the earlier strategy of giving participants more voice in the research process beyond just being sources of facts, knowledge and experiences.
Beyond the reflexive acknowledgement of the mediating role of the researcher in the construction of interview data, is the strategy to acknowledge the various actants or inscription devices that shape this construction. Latour (2005) suggests that non-humans can be more than placeholders in the social world but rather, can play active roles in the assemblage of social realities. Various actants appear present in typical Grounded Theory interviewing such as: audio-recording devices, transcription text, dominant discourses, coding techniques, local settings, and internet mediated interaction.
It is difficult to discern what mediating role these, and other, actants play within given Grounded Theory studies. To attempt to follow the traces left by each might be a never-ending task. However, it still appears important to allow for some transparency of the actants that might be involved; at the very least, to acknowledge their existence and the mediating role they might have in the construction and development of
interview data.
The result of such a strategy would be a further move away from the original positivistic orientations of the Grounded Theory approach. The concern would no longer be with hiding the mediating role of the researcher, and the tools and technologies of his or her trade, but rather to acknowledge the important role these play in the construction of interview data. Essentially, this strategy holds that forthrightness will replace the illusion of objectivity. As such, in the interest of transparency over illusionary objectivity, efforts were made in our study of marijuana users to highlight the various actants which might have played a mediating role in the construction and development of interview data such as: the audio recorder, the computer, textual transcriptions, and the location of the interview.
The final limitation outlined, that interview data is often treated as a mirror of the lived experiences of respondents, is the most challenging to reconcile. The real challenge appears to be in preserving some of the lived experiences of the respondents within their representations. This is a difficult task with all of the mediating factors that go into interview data construction, from the questions that the interviewer asks to the medium through which the interview is conducted. One possible strategy is to try to avoid what Institutional Ethnographers refer to as "institutional capture" (Smith 2005, p. 119). Institutional capture occurs when both the interviewer and the respondent are familiar with a given institutional discourse. The interview is then conducted in a manner whereby the interactions are informed by this discourse, which can lead to the overlooking or pushing aside of lived experiences.
Thus, one strategy to maintain a measure of the lived experience of the respondents would be to avoid assuming that certain forms of
knowledge are shared in common. Instead, a task of the researcher and respondent might be to develop new meanings of taken-for-granted terms, as DeVault (1999) suggests. Using a basic example from the marijuana study, many of the marijuana users suggest that one of their main reasons to use marijuana is to get 'high'. This seems like a pretty self-evident and straightforward claim. But, we might assume a shared understanding of the term 'high', instead of discussing what the lived experience of being 'high' entails. Regardless of the researchers own understandings and experiences of being 'high', the object of inquiry should be how the participant experiences being 'high'.
Another strategy to maintain some of the complexity of the lived experiences of respondents, rather than tapering them through a series of grounded theory processes, might be to give respondents more responsibility in the processes of theorization. Latour (2005) writes: "You have to grant them back the ability to make up their own theories of what the social is made of. Your task is no longer to impose some order, to limit the range of acceptable entities, to teach actors what they are" (p. 11/12). Just as it might be a useful strategy to have participants label and classify themselves, as well as to discuss the formulation of research questions, it might be useful to have increased participant involvement in the development of concepts and theories.
A method of doing this might be to discuss emerging ideas and theories with respondents in interviews of collaborative knowledge construction as previously described. Many of the theoretical ideas that have been published out of our marijuana study have been shaped by discussions with our participants. Another method could involve having respondents who appear particularly knowledgeable about a particular field of inquiry look over the developing concepts and theorizations as they emerge. This approach was used by Bob Stebbins (1987), who had
professional football players look over the findings that he generated to confirm that his concepts and theorizations fit with the experiences of professional football players in Canada. A more in-depth approach could be done through collaborative research, whereby one of the primary researchers acts as a participant or member of a group under study. An example of this is the collaboration of Dragu and Harrison (1989) in a study of the female strip trade; Margaret Dragu worked as a stripper before the study.
Conclusion
A number of limitations of traditional Grounded Theory interviewing approaches have been identified. These include: the labelling and categorizing of participants on their behalf, the treatment of participants as vessels of fact and knowledge, the lack of reflexivity on the mediating role of the researcher and the technologies he or she employs, as well as the use of data as reports that are believed to mirror the lived experiences of the respondent. Each of these limitations appears to arise from outdated Grounded Theory interviewing methods that still seem to be developed in attempts to gain the legitimacy of the natural sciences. The problem is that these approaches have been based on false ideals. Scientific fact is not created external to scientists, and the various technologies and inscription devices they use. To hide these mediators might actually move the researcher further from an objective position, rather than closer to one.
Taking these criticisms seriously, Grounded Theory interview strategies could involve moving away from this illusion of objectivity to acknowledge the role of the researcher, and various other mediators, in the construction of interview data. It might also involve allowing interview respondents more responsibility in the construction of
interview data and its subsequent analysis. The ultimate strategy could be to have respondents themselves tell the researcher who they are and how they organize their realities, rather than have the Grounded Theorist do this for them. By revoking its aspirations to mirror the natural sciences, instead opting for increased reflexivity, Grounded Theory could become better suited to study the contemporary social world.
References
Berg, B. L. (2004). Qualitative research methods for the social sciences, 5th Edition. Toronto: Pearson.
Bryant, A. (2003). A constructive/ist response to Glaser. Forum Qualitative Sozialforshung/Forum: Social Research, 4 (1), available online at: www.qualitative-research.net.
Bryman, A. (1988). Quantity and quality in social research. London: Unwin Hyman.
Charmaz, K. (2000a). "Qualitative interviewing and grounded theory analysis." In Handbook of interview research: Context and method, J. F. Gubrium & J. A. Holstein (eds.). Thousand Oaks, CA: Sage. Pp. 675-694.
Charmaz, K. (2000b). "Grounded Theory: Objectivist and Constructivist Methods." In Handbook of Qualitative Research, N. K. Denzin & Y. S. Lincoln. Thousand Oaks, CA: Sage. Pp. 509-539.
Devault, M. L. (1999). Liberating method: Feminism in social research. Philadelphia, PA: Temple University Press.
DeVault, M. L. (1991). Feeding the family: The social organizing of caring as gendered work. Chicago: University of Chicago Press.
Dragu, M., & Harrison, A. (1989). Revelations: Essays on striptease and sexuality. Gibsons Landing, BC: Nightwood Editions.
Douglas, J. (1984). Creative interviewing. Thousand Oaks, CA: Sage.
Glaser, B. (1992). Basics of grounded theory analysis: Emergence vs. forcing. Mill Valley, CA: Sociology Press.
Glaser, B. (1978). Theoretical sensitivity: Advances in the methodology of grounded theory. Mill Valley, CA: Sociology Press.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative research. London: AldineTransaction.
Karp, D. A. (1994). Living with depression: Illness and identity turning points. Qualitative Health Research, 4 (1): 6-30.
Latour, B. (2007). "From associations to modes of existence." Presented at the Annual Meeting of the British Sociological Association, London, April 13.
Latour, B. (2005). Reassembling the social. New York: Oxford University Press.
Latour (1999). Pandora's hope: Essays on the reality of science studies. Cambridge, Mass.: Harvard University Press.
Latour, B. (1996). Aramis or the love of technology. Cambridge, Mass.: Harvard University Press.
Latour, B. (1988). The Pasteurization of France. Cambridge, Mass.: Harvard University Press.
Latour, B. (1987). Science in action: How to follow scientists and engineers throughout society. Cambridge, Mass.: Harvard University Press.
Latour, B., & Woolgar, S. (1986). Laboratory life: The construction of scientific facts. Princeton: Princeton University Press.
Law, J. (2004). After method: Mess in social science research. New York: Routledge.
Lee, R. M. (2004). Recording technologies and the interview in sociology, 19202000. Sociology, 38(5): 869-889.
Mason, J. (2002). "Qualitative interviewing: Asking, listening and interpreting." In Qualitative Research in Action, T. May (ed.). London: Sage. Pp. 225-241.
Osborne, G., & Fogel, C. (2008). Understanding the motivations for recreational marijuana use among adult Canadians. Substance Use and Misuse, 43(3), 539572.
Osborne, G., & Fogel, C. (2007). The normalization of marijuana use by adult Canadian users. International Journal of Crime, Criminal Justice, and Law, 2(2), 201-225.
Pence, E. (1996). "Safety for battered women in a textually mediated legal system." Dissertation: University of Toronto. Supervisor: D. E. Smith.
Silverman, D. (1993). Interpreting qualitative data: Methods for analysing talk, text and interaction. London: Sage.
Smith, D. E. (2005). Institutional ethnography: A sociology for people. Toronto:
AltaMira Press.
Smith, D. E. (2003). "Resisting institutional capture: A research practice." In Our Studies, Our Selves, B. Glassner & R. Hertz (eds.). New York: Oxford University Press. Pp. 150-161.
Stebbins, R. A. (1987). Canadian football: The view from the helmet. London, On: CSHS.
Strauss, A. L. (1987). Qualitative analysis for social scientists. New York: Cambridge University Press.
Strauss, A. L., & Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory. Thousand Oaks, CA: Sage.
Widdicombe, S. (1998). "'But you don't class yourself': The interactional management of category membership and non-membership." In Identities in Talk, C. Antaki & S. Widdicombe (eds.). London: Sage. Pp. 52-70.
Woolgar, S. (1988). Knowledge and reflexivity: new frontiers in the sociology of knowledge.Thousand Oaks, CA: Sage.