• Tidak ada hasil yang ditemukan

C ONVENIENCE S AMPLE

A convenience sample can be defined as a sample in which research participants are selected based on their ease of availability. Essentially, individuals who are the most ready, willing, and able to participate in the study are the ones who are selected to participate. In qualita- tive research, it may be helpful to use a convenience sample to test the appropriateness of interview ques- tions in an inexpensive and quick way by approaching an interested group of people first before embarking on a larger, longer, and more expensive study.

One common example of convenience sampling is found in psychology, where introductory psychology

students are frequently recruited to answer psycholog- ically oriented research questions, such as their moti- vation to exercise, as part of their introductory coursework in psychological research. In essence, these students have been asked to answer questions about their motivation to exercise because they are readily accessible. That is, the researcher could recruit them directly from introductory psychology classes and did not need to venture into the wider community to conduct the research.

Although this type of sampling technique can most assuredly save the researcher time and money at the recruitment stage, it is not without its drawbacks. If the researcher recruits students from a psychology class at a particular academic institution, for example, it is difficult to know whether or not the students’

motivation to exercise is reflective of motivation to exercise in other contexts. Similarly, it may be that students are perhaps less motivated to exercise than members of the larger society because they are too busy with their coursework. Hence, it is difficult to assess whether or not the study’s findings regarding motivation to exercise can apply to students and the population at large. For this reason, convenience sam- ples can lack transferability (or external validity) in qualitative research. In quantitative projects, it can be difficult to generalize the results beyond the original sample from which the data were collected. However, because qualitative researchers are typically interested in studying specific groups of people rather than gen- eralizing to larger populations, it may be tempting for researchers to view this issue as a less significant problem than it is for quantitative researchers.

However, it is still important to be aware that the par- ticipants recruited are not necessarily reflective of the population being studied. For example, when recruit- ing for a study about the views of breast-feeding mothers, it is possible that those who answer the advertisement and come to talk to the researchers are the ones with the strongest opinions. They may be individuals who have the most vested interest in shar- ing their stories because they want to effect change in society’s attitudes toward breast-feeding. With this in mind, qualitative researchers should be aware that the people who are recruited most readily are not neces- sarily reflective of all viewpoints.

In sum, a convenience sample can be described as a group of participants who have been recruited for a given study because they were readily accessible.

Using a convenience sample can be both time- and 124———Convenience Sample

cost-effective, but caution about the nature of the results should be considered in their interpretation.

Kristie Saumure and Lisa M. Given

See alsoNonprobability Sampling; Purposive Sampling;

Sampling; Snowball Sampling

Further Readings

Henry, G. T. (1990).Practical sampling.Newbury Park, CA: Sage.

Patton, M. (2002).Qualitative research and evaluation methods(3rd ed.). Thousand Oaks, CA: Sage.

C

ONVERGENT

I

NTERVIEWING

Convergent interviewing (CI) is a technique developed by Bob Dick from the University of Queensland in Australia. CI aims to collect, analyze, and interpret people’s experiences, opinions, attitudes, beliefs, and knowledge that converge around a set of interviews. It was created primarily to address issues in underre- searched areas. Although used and documented mainly by Australian researchers in marketing/business disci- plines with foci on organizational change and develop- ment, it has been extended and adapted to broader social science and health research. This entry describes the CI process and compares and contrasts it with other qualitative techniques and methods. CI provides a mechanism to flexibly structure research projects while using unstructured content to enable greater reflexivity throughout the different phases of research.

CI permits in-depth interviewing by promoting a cyclical research process that requires ongoing analy- sis as part of the overall strategy. It is most suitably applied when multiple interviewers are being used in a project, and its aim is to document priority issues when these converge over a series of interviews.

According to David Carson and colleagues, this process is iterative and, thus, enables continuous refinement. Interviewers engage in a constant compar- ative reflexive process that permits detailed rich con- tent and theoretical sampling as researchers seek to continuously test emerging interpretations from early interviews in subsequent interviews.

Although designed to accommodate the use of multiple interviewers, a single interviewer can be used

provided that he or she can hold discussions with one or more members of the research team. CI provides a structured process while using unstructured content to enable greater reflexivity throughout the different phases of research.

Several aspects of CI bear mentioning, including use of prior literature, sampling, designing the inter- view questions, data capture, and comparing and con- trasting during the interview process through to the analysis phases.

Prior Literature

Unlike with grounded theory approaches and earlier articulations of CI, Bob Dick encouraged researchers to engage with the literature from the beginning. Prior knowledge is designed to facilitate the development of rapport (interviewers/researchers can make more convincing “sounds” showing their understanding of what participants are expressing), to enable researchers to recognize potential priority issues (aris- ing out of discussions with participants as well as the literature), to aid in the development of a relevantly worded opening question (which can keep partici- pants talking without needing to ask additional ques- tions), to assist in the selection of an appropriate sample, and to increase researchers’ confidence when conducting interviews. As Carson and colleagues noted, engaging with the literature throughout the process—both data collection and data analysis—

permits an “unfolding” of the literature as priority issues emerge from the interviews.

Sampling and

Developing Interview Questions Sampling is heterogeneous, seeking maximum varia- tion that may be augmented through snowball sam- pling strategies as deemed appropriate. Participants are sampled as information-rich cases where each subse- quent interview is designed to pursue areas of agree- ment or disagreement on what the priority issues are for the phenomenon under study. To accomplish this, a very general opening question is used to guide the process. This opening question is designed to have the participant speak for a long period of time (up to an hour) without the interviewer needing to ask additional questions beyond gentle follow-ups on what is raised specifically by the respondent; for example, when the Convergent Interviewing———125

respondent discusses issues of trust, the interviewer might say, “Trust? Please elaborate.” Dick argued that this type of approach helps to ensure that the inter- viewer does not guide or introduce the content.

The development of the opening interview ques- tion requires the participant to comment on, or share a story about, both the strengths and weaknesses, or what was “good” and “bad,” about the phenomenon under study. Although seemingly similar to the critical incident technique, the CI process differs in funda- mental ways. Unlike with the critical incident tech- nique, its interviewers do not present participants with a brief statement for their comment on the activity, nor do they query how observed behavior is to be evalu- ated and classified. Moreover, the CI process does not involve the interview guide being developed prior to data collection (beyond the general opening question), nor are interview questions direct and typically sequential as they tend to be in critical incident inter- viewing techniques. Rather, the interview guide is developed and built on over time using topics raised during earlier interviews in subsequent interviews as convergence over priority issues is sought through constant comparison.

Data Capture

Dick originally preferred that researchers adopt a self- designed memory system of 20 keywords raised in the interview to reconstruct the key ideas from the inter- view rather than rely on audiorecordings. However, alternative approaches using CI that have evolved

encourage the use of audiorecordings, and Dick even softened his stance on this point. Like grounded theory approaches, CI encourages the interviewer- researchers to faithfully produce short notes and field- notes in addition to making audiorecordings of interviews for later transcription. Fieldnotes also facil- itate the discussions of priority issues emerging from the interviews between interviewers.

Comparing and Contrasting

Dick designed CI to accommodate multiple interview- ers; although this is not always feasible in academic projects due to limited resources. In many projects, the use of multiple interviewers can introduce sub- stantial variation in quality and reliability that can be minimized if interviewers are given prior training and use a standardized interview guide. However, in CI the ideal circumstance is to have multiple interviewers working in pairs but conducting their own individual interviews. This process involves the interviewers meeting following each pair of interviews to discuss what ideas or issues were raised by participants. They check for early convergence and develop questions to be asked only after the original opening question has been exhausted to verify areas of agreement or dis- agreement. With each discussion among the inter- viewers, more prompts are developed over the course of data collection but often are not asked because par- ticipants themselves raise these areas unprompted by the researcher. These frequent discussions among the interviewers are designed to ensure that areas of 126———Convergent Interviewing

Table 1 An Example Matrix of Agreements and Disagreements About Issues in Five Convergent Interviews Issue

Respondent 1 2 3 4 5 6 7

A Yes Yes – – – – –

B Agree Disagree Yes Yes – – –

C Agree Disagree Agree Agree Yes Yes –

D Agree Disagree Agree Agree Agree – Yes

E Agree Agree Agree Disagree Agree Agree Agree

Source:Carson, D., Gilmore, A., Gronhaug, K., & Perry, C. (2001).Qualitative research in marketing.London: Sage. Reprinted with permission by Sage Publications, Ltd.

Notes:This table is for illustrative purposes only. To illustrate, Respondent A raises two issues during the interview. Respondent B agrees with Issue 1 but disagrees with Issue 2. Respondent B also raises Issues 3 and 4. Respondent C agrees with Issues 1, 3, and 4 but disagrees with Issue 2 and raises Issues 5 and 6 and so forth.

agreement and disagreement are consistent across the sample and that no new priority areas or disagree- ments are being introduced.

Frequent discussions among interviewers also facilitate preliminary and detailed analysis. Unlike with grounded theory, where early analysis can lead researchers in unanticipated directions or necessitate greater sampling, the inherent process of CI forces interviewers to identify differences as the interviews progress. By capitalizing on the use of multiple inter- viewers, it is possible to reach saturation or conver- gence on priority issues more quickly because interviewers can gain from the experiences of other participants that they did not personally interview.

These iterative discussions between the interviewers may also ensure that the epistemology (theories of knowledge or how knowledge is demonstrated) and ontology (examples of social reality) of the research project are maintained throughout the process of data collection and data analysis. These discussions can also overcome obstacles that may arise when dealing with large multidisciplinary research teams. S. Michelle Driedger and colleagues presented a more detailed discussion of how CI facilitates this by making explicit project epistemologies and ontologies.

Table 1 illustrates the CI process.

S. Michelle Driedger

See alsoConstant Comparison; Critical Incident Technique;

Data Saturation; Grounded Theory; Interviewing;

Reflexivity

Further Readings

Carson, D., Gilmore, A., Gronhaug, K., & Perry, C. (2001).

Qualitative research in marketing.London: Sage.

Charmaz, K. (2000). Qualitative interviewing and grounded theory analysis. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research(2nd ed., pp. 675–694).

Thousand Oaks, CA: Sage.

Dick, B. (1990).Convergent interviewing.Brisbane, Australia: Interchange.

Dick, B. (1998).Convergent interviewing: A technique for qualitative data collection.[Electronic version]. Available from http://www.uq.net.au/action_research/arp/iview.html Driedger, S. M., Gallois, C., Sanders, C. B., & Santesso, N.,

on behalf of the Effective Consumer Investigator Group.

(2006). Finding common ground in team-based qualitative research using the convergent interviewing method.

Qualitative Health Research, 16,1145–1157.

Patton, M. Q. (1990).Qualitative evaluation and research methods.Newbury Park, CA: Sage.

Rao, S., & Perry, C. (2003). Convergent interviewing to build a theory in under-researched areas: Principles and an example investigation of internet usage in inter-firm relationships.Qualitative Market Research: An International Journal, 6,236–247.

Riege, A. M., & Nair, G. (2004). The diversity of convergent interviewing: Applications for early researchers and postgraduate students.Marketing Review, 4,73–85.

C

ONVERSATIONAL

I

NTERVIEWING

Conversational interviewing is an approach used by research interviewers to generate verbal data through talking about specified topics with research partici- pants in an informal and conversational way.

Although all qualitative interviewing relies on speak- ers’ everyday conversational resources, conversa- tional interviewing foregrounds aspects of sociability, reciprocity, and symmetry in turn taking found in mundane conversation.

Whether research interviews are structured, semi- structured, or unstructured, interviewers and intervie- wees rely on taken-for-granted assumptions about how everyday talk occurs and how speakers make meaning of one another’s utterances. In emphasizing features of mundane conversation, conversational interviewers strive to facilitate a research environ- ment in which participants feel free to participate in extended discussions of research topics in a less hier- archical environment than that convened in structured interview settings.

Conversational interviews have long been used by anthropologists and sociologists to talk to people for the purpose of generating data in field studies and ethnographic work. Although this form of interview- ing is used by ethnographic researchers undertaking prolonged fieldwork, it is also popular among qualita- tive researchers who use open-ended, in-depth, or unstructured interview formats and among researchers who advocate feminist and emancipatory approaches to research interviewing.

Methodological discussions of qualitative inter- viewing have frequently associated conversation with research interviews. Regardless of the theoretical ori- entation taken by different methodologists, the notion that conversation is synonymous with interview is widespread and the qualitative interview has been Conversational Interviewing———127

described variously as a “guided conversation,” a

“conversation with a purpose,” a “professional con- versation,” and a “directed conversation.” Although ordinary conversation is the bedrock on which inter- view interaction relies, there are distinct differences between conversational talk and interview interaction.

Contrasting Conversation and Research Interviews

Research interviews are frequently conducted with strangers, and researchers must first arrange a time and place for talking with research participants—either in person or via telephone. The scheduling involved in research interviews is unlike the haphazardness inherent in everyday conversation. In making use of conversational interviews in prolonged fieldwork, ethnographic interviewers are better able to emulate the spontaneity of conversation in their interviewing prac- tice when they pose casual questions to participants about what is going on as part of their participant obser- vations. Qualitative researchers must abide by institu- tional procedures for informed consent, and the requirements for obtaining written or oral consent from participants for their participation in research also devi- ate from everyday conversation. Thus, conversational interviewers must work against these formal constraints by simultaneously orienting participants to the purpose of upcoming interaction and setting an informal and casual tone for extended conversation.

Conversations routinely take place between people who are known to one another; thus, rapport building is not necessarily facilitated in the talkprior toa con- versation but might be thought of as being producedby good conversation. In everyday life, initiating conver- sations with strangers is a delicate task, and topics must be introduced judiciously by speakers if the inter- action is to be prolonged. Just as initiating conversa- tions with strangers is delicate work, conversational interviews with strangers must be handled with sensi- tivity, and talk leading up to the discussion of research topics is thought to be important for rapport building.

Thus, conversational interviews require that at the out- set of interviews, researchers facilitate the kind of small talk familiar to conversationalists who have just met; for example, in Western societies, this could include observations concerning travel, weather, or occupations. Conversational entrées to research inter- views are seen to facilitate openness, informality, and rapport between interviewers and interviewees.

Viewed superficially, everyday conversation seems chaotic and unfocused, with speakers collaboratively involved in asking and responding to questions in an interactive sequence that involves new topics and continuous clarification of speakers’ meanings.

Conversation analysts have shown mundane conversa- tion to be complex, with routine sequences that speak- ers deliberately use as resources to accomplish daily activities such as greetings, making excuses, apolo- gizing, complaining, and closing conversations. In research interviews, the central activity is generat- ing data via question-and-answer sequences, and researchers set the scene for subsequent interview talk by providing participants with an outline of the research topic. Interviewers pose questions and follow up speakers’ answers with requests or probes for fur- ther explanation. Conversational interviewers strive to create a friendly and informal atmosphere in which participants are respected as equal partners who are free to share their understandings concerning the research topic.

Everyday conversations generally do not have a preset topical agenda for talk; in contrast, researchers use questions drawn from a semi-structured interview protocol or a list of prepared topics to introduce top- ics of interest into conversational interviews. Whereas in structured interviews there is little opportunity for interviewees to introduce new topics or ask questions, conversational interviewers are open to new directions in the talk provided by participants and are likely to respond in an open and authentic way to questions that interviewees might pose to them.

In everyday conversation, speakers do not neces- sarily unpack statements made by others in systematic ways. Consider, for example, the following exchange:

Speaker A: How’s work?

Speaker B: It’s going well.

In a conversation with a friend, Speaker A would be unlikely to reply to Speaker B with, “Tell me more about what is going well for you at work,” and without other nonverbal cues from Speaker B, Speaker A would likely move to a new topic of talk. Conversational inter- viewers, however, unpack the “glosses” provided by interviewees. Thus, in response to the same question in an interview setting, Speaker A’s response to Speaker B’s utterance might not be out of place. Because the utterances made by both speakers in a conversational interview are produced for research purposes, the 128———Conversational Interviewing

interviewer is more likely to take the roles of topic ini- tiator, question poser, and clarification seeker than is the interviewee. The interviewer’s ability to pose ques- tions, seek further explanation, and initiate topics as part of his or her research agenda, then, tends to pro- duce a more asymmetrical relationship than one might see in ordinary conversation between equals.

Finally, in everyday conversations, it would be unusual for interaction to be recorded. In conversational interviews, with the prior consent of participants, researchers will make written records of participants’

utterances that will become data for analysis and inter- pretation, and excerpts will be used for research reports.

Whereas early qualitative researchers frequently relied on handwritten notes from research interviews that were expanded at a later point in time, qualitative researchers today commonly audio- or videorecord interviews for transcription, analysis, and interpretation.

Issues in Using

Conversational Interviews

Although conversational interviewing as a format for eliciting data for research projects has become increasingly favored by qualitative researchers, there are both benefits and limitations to this approach to data generation. A friendly and skilled interviewer facilitates an in-depth exchange with research partic- ipants through the use of conversational interaction.

Conversational strategies include incorporating infor- mal talk, showing flexibility in allowing topic shifts and questions from interviewees, inviting reciprocity by openly responding to questions and comments from interviewees, and treating conversational part- ners sociably—with respect, care, and intensive lis- tening. In response to this framing of the interview by the researcher, interviewees may provide confes- sional and self-revealing details about their lived experiences, beliefs, and perceptions. Although some methodologists have referred to data generated in such exchanges as more authentic than those derived in more structured formats, others have critiqued this view of interviewing as naive and simplistic, instead emphasizing the manipulative potential of conversa- tional interviewing. In generating disclosure from their participants via casual, friendly, and informal interview formats, researchers may be accused of manipulating their participants for personal gain.

Furthermore, data generated via conversation pro- vide much potential for manipulation by researchers

as they code, analyze, interpret, and represent speak- ers’ words. Researchers using conversational inter- views in their work must address a range of questions concerning data generation and representation: When is it appropriate for a researcher to contribute personal accounts and views to the interaction? What are the implications of a researcher’s contributions to the talk for what participants say next? Given that speakers’

talk routinely includes slips and repairs, what features of talk should be transcribed and how should talk be edited for final reports? How much of a researcher’s contribution to the generation of the talk should be included in reports? What means of analysis should be used for conversational talk in which both speakers contribute to equal degrees?

Given that interviewing as a method of data genera- tion is complex work, relying on the conversational skills of both interviewer and interviewee, the answers to these questions are likely to be different for each and every researcher in each and every interview encounter.

Kathryn J. Roulston

See alsoFocus Groups; In-Depth Interview; Probes and Probing; Semi-Structured Interview; Structured Interview;

Unstructured Interview

Further Readings

Fontana, A., & Frey, J. H. (2005). The interview: From neutral stance to political involvement. In N. K. Denzin &

Y. S. Lincoln (Eds.),The SAGE handbook of qualitative research(3rd ed., pp. 695–727). Thousand Oaks, CA: Sage.

Kvale, S. (1996).InterViews: An introduction to qualitative research interviewing.Thousand Oaks, CA: Sage.

Kvale, S. (2006). Dominance through interviews and dialogues.Qualitative Inquiry, 12,480–500.

Rubin, H. J., & Rubin, I. S. (2005).Qualitative interviewing:

The art of hearing data(2nd ed.). Thousand Oaks, CA: Sage.

Spradley, J. (1979).The ethnographic interview.New York:

Harcourt Brace Jovanovich.

C

ONVERSATION

A

NALYSIS

Conversation analysis (CA) has become the established label for a quite specific approach to the analysis of interaction that emerged during the 1960s in the work Conversation Analysis———129