Chapter 5: Qualitative Data (Part 2)

Qualitative Data Gathering Research Designs

In their search for understanding communication phenomena, researchers have multiple qualitative methods from which to choose. Depending on a variety of factors (such as the nature of the research question, access to participants, time and resource commitments, etc.), researchers may select one or more of the following methods:

  • ethnography
  • in-depth field interviews
  • focus group interviews
  • the collection of narratives. 

Selecting the appropriate method for data collection is a vital component of the research process. Regardless of the method selected, researchers must reconcile the established traditions of the methodology with the specific requirements of the group or individuals participating in the research. We now discuss how to plan and implement your qualitative study. This section begins with topic selection and research focus and then proceeds to a discussion of each of the different qualitative methods.

Identifying the Research Setting, Research Group, and Research Focus

Researchers have several choices when deciding how to proceed with a qualitative study. Some studies may begin with a specific communication concept, such as family communication. Researchers then begin to identify potential study participants. On other occasions, a researcher might be interested in a specific setting, such as a tattoo parlor. Gaining access to that setting to see what interesting communication concepts emerge would be very helpful. In both cases, research proceeds inductively, and conclusions emerge from the carefully gathered data. Regardless of how the initial inspiration strikes, the subsequent steps in the procedure follow a similar pattern. The following chart demonstrates the typical qualitative data gathering process:

Data Gathering Process

Selecting a topic and narrowing the research focus. During the earliest phases of the qualitative research process, researchers are tasked with identifying a focus for their study. Like all research, the individuals conducting the study are often drawn to those communication phenomena that are of the most interest to them. Perhaps a researcher has a friend or a family member who recently met his or her spouse through an on-line dating service and the researcher becomes interested in understanding how on-line dating develops. An initial research question might be, “What are the normative behaviors regarding on-line courtship?” From this point, it is important for the researcher to develop a rationale for the research. Sometimes, as in the case of on-line dating, the rationale is self-evident. As the number of people who participate in on-line dating continues to grow, it becomes an important and useful social activity to investigate. Regardless of whether or not the utility of the study seems self-evident, the researcher has an obligation to demonstrate the relevance of his or her study rationale through a review of the existing literature.

The literature review is an important component of any carefully designed research study. Although existing theory typically guides the more deductive approach of quantitative research, there are several differences regarding the more inductive, qualitative literature review. In qualitative research, the literature review is not completely finalized before data collection begins. In fact, in many cases, the literature review proceeds alongside the interviews or observations in which the researcher may be engaged. In the previous example regarding on-line courtship, the researcher would likely construct a literature review based on articles that examine on-line dating behaviors, as well as articles that examine off-line dating behaviors. However, if during the process of interviewing participants, several interviewees discuss the importance of having friends who accepted and encouraged their on-line dating attempts, the importance of a concept like social support might emerge. The qualitative researcher, well into the process of interviewing, might gather existing literature on social support and then ask questions regarding social support in future interviews. In some cases, the literature review continues to grow and develop as the data is being collected.

Another difference, though to a lesser degree, is that qualitative researchers often conduct a broad, rather than a deep, literature review—at least initially. The broad approach familiarizes the researcher with multiple topics that seem related to his or her research purpose. However, given that the specific data, rather than general theory, guides the entire qualitative process, the researcher can choose to analyze more deeply those topics that begin to emerge from the interaction with, or observation of, the participants. It is imperative that researchers are responsive to data collected over the course of the qualitative project so that they can augment or de-emphasize segments of the literature review as needed.

Choosing the appropriate methodology and accessing the setting or participants. Although you can choose from many acceptable qualitative methods, including case study analysis, autoethnography, or qualitative content analysis, this chapter will focus on four common methods of collecting qualitative data: ethnography, interviewing, focus group interviewing, and narrative inquiry. Each of these methods, including their strengths and limitations, will be discussed in more detail later in this section. Choosing the appropriate method is based on a variety of factors including the nature and scope of your research question or questions, access to study participants, and researcher training and familiarity with potential methods, to name a few. As with any type of scholarly research, the law of the hammer need not apply. The law of the hammer states that if the only tool available is a hammer, then every problem will resemble a nail. If a researcher is trained and comfortable with conducting focus groups, but the best method of data collection for answering a specific research question is ethnographic research, then the researcher needs to take the necessary amount of time familiarizing himself or herself with the steps of an ethnography rather than forcing the question to conform to a focus group format. As always, the research purpose should guide the methodology, rather than the methodology guiding the purpose of the research.  

Access and trust are fundamental elements of successful data collection in qualitative research. A researcher must be able to access the research setting or interview participants in order to gather data. If the researcher is unable to gain access, it is possible that the study will have to be abandoned or significantly altered. In the case of Tom Hall’s research into government secrecy and its effects on democracy several years ago, he attempted to interview the members of the U.S. Senate and the House of Representatives who served on the Senate and House intelligence committees, respectively. Of the nearly thirty members serving at the time (2003-2004), he was not able to gain access to a single congressperson. In fact, only a handful responded to his numerous attempts at contact. Due to this lack of access, he had to shift his project entirely to a textual analysis of public documents, such as Executive Orders, Congressional Research Reports, and Department of Energy documents, in order to proceed with the research. This lack of interview access changed the general purpose of the research project and significantly altered the research questions he sought to answer. Needless to say, flexibility is an important attribute for a qualitative researcher.

In many cases, whether or not a group or organization provides access stems from whether or not the members of that organization trust the researcher. Imagine your suspicion if an individual appeared at your place of employment one day asking questions and writing information in a notebook. Just because a supervisor or gatekeeper has granted access to a researcher does not mean that all of the members of the organization or group are ready to trust the motivations of the researcher. Trust is person specific. The key to collecting solid data is for those individuals who are being observed or interviewed to understand why you are there and how you plan to use the data. If an interviewee does not trust your motivations, it is highly unlikely that s/he will be forthcoming in her/his responses—if s/he chooses to participate at all. While access to a group often relies on gaining the permission of a gatekeeper or other organizational leader, trust develops over time.

Examples of the early stages of qualitative research. Sally is interested in studying a successful student organization on her campus. She realizes that PRSSA is an award-winning public relations organization on campus, so she contacts the faculty advisor and the current student leaders in the organization. Sally must gain the consent of the faculty advisor and other student leaders in order to begin her research on PRSSA. Sally decides to immerse herself as fully as possible with PRSSA—attending meetings, interviewing members, and observing committees, among other activities. Over the course of her research, all of Sally’s data collection efforts will be focused on this one organization. Although the initial approval of the faculty advisor and the student leaders of the organization are vital to gain access, it is important to remember that Sally will still need to gain consent from any other organization members who participate in her study.

Another example involves Bob who is also interested in studying reasons for participation in student organizations on his campus. Rather than focusing on a single organization, he decides to interview multiple participants in multiple organizations. Over the course of several weeks or months, Bob interviews numerous individuals and learns why they chose to involve themselves in student organizations. Bob uses a snowball or volunteer sample in order to recruit participants for his study. While there is no single faculty advisor or organization leader needed to gain access to these groups, Bob will still have to acquire individual consent from all of the participants in his study.

The early stages of a qualitative project are crucial for providing the foundations for a credible study. Developing the research purpose, examining the existing literature, selecting an appropriate method, and gaining access to and the trust of participants are all necessary steps.  

Qualitative Data Collection

In this section the authors discuss each of the more common qualitative methods, including the purpose, the steps involved in the particular method, and the strengths and limitations of the method. An extended example of each method is provided before moving on to the next method. The methods discussed include: ethnography, interviewing, focus group interviewing, and narrative interviewing.

Ethnography. This method is the total immersion of the researcher into the research setting. The roots of ethnography lie in cultural anthropology, which is when researchers attempt to fully understand the culture of a group by integrating themselves into the culture under investigation. Some communication scholars, such as Lawrence Wieder (1999), consider ethnography to be the main qualitative method. Although observation is the central practice of ethnographic research, a researcher may employ multiple other qualitative methodologies during the course of the project. Ethnography is the method of choice when a researcher decides to study the participants within their natural environment. Therefore, a study examining the communication patterns of college wrestlers would involve extensive interaction with a college wrestling team.

Traditional ethnography concerns a researcher or group of researchers studying the activities of a specific group or culture Over the last twenty years, an additional type of ethnographic research has emerged, where a person tells the story of some experience within their own life, with a scholarly purpose. This practice of autoethnography has become more popular in the past few years, and its validation as a legitimate form of knowledge development has also increased. Autoethnography places the researcher at the center of the investigation by directing the analysis towards the researcher’s role in the natural setting. By definition, an autoethnographer is a complete participant in his or her research, and is engaged in a full scale, in-depth, critical analysis on his or her life as it is being lived. A college athlete might detail his or her experiences as a member of a team—critically exploring his or her assimilation and identification with the group or chronicling the personal difficulties of competing at the highest level athletically. Whatever the specific focus of research, the data is drawn from detailed accounts of a person’s own experiences.  

The purpose of ethnographic research. Given though the individual goals vary from one ethnographic project to another, the overall purpose of the research remains the same: to accurately capture the social activities of the group or organization being studied. Frequently, the researcher does not have clearly defined research questions, preferring instead to capture, through observation and interviews, the social practices of the organization members. The purpose is to experience the participants in their natural setting and interpret those experiences accurately, in order to develop a better understanding of the interaction processes of the social group. Remember that one of the overarching goals of qualitative research is developing an understanding of an event, phenomenon, communicative practice, or cultural group. Ethnography is one of the ways that a researcher might accomplish this goal.

The steps in the ethnographic process. You should follow four distinct steps when conducting ethnographic research.

  1. Identify the research site. In ethnographic research, the principle investigator is often attracted to a particular group or organization. The person conducting the study may already be a member of the group (emic) he or she desires to research. An example of this would be when an individual who is a member of a book club decides to research the group in order to understand the group better. An etic approach might involve a graduate student who is interested in high school coaching selecting an athletic team to observe. Observation is a necessary component of ethnographic research. The full range of observation techniques are available to the researcher—from complete participant to complete observer.
  2. Gain access to the organization. Gaining access to an organization is vital to ethnography. Simply put, without access, ethnography is impossible. In order for the research to proceed, investigators must gain access to the organization or group. Gaining access relies on building trust with gatekeepers, informants, and other key organizational personnel. For example, if a researcher wanted to study the cultural climate of a Communication Studies Department, he or she might first begin by approaching the head of the department for approval. In this example, the department head might act as a gatekeeper—permitting or denying the research. However, even with the consent and trust of the department head, successful data collection and a successful research project are not assured. The researcher would likely need to approach each of the professors and staff in order to gain their consent to be observed during meetings and classes and office activities. While the consent of the department head may be enough to gain access to the data site, the success of the data collection portion of the project relies on earning the trust of as many of the organizational members as possible. 
  3. Collect and analyze data. During this stage, the real work begins. Researchers immerse themselves in the natural environment of the participants, collecting copious field notes and analyzing the data frequently. This is by far the most time consuming portion of the research process. At a minimum, most ethnographers spend six months engaged in this portion of the research and spending one or two years immersed in a group is not uncommon.
  4. Leave the field of investigation. When subsequent observations are no longer producing new data, the researcher is ready to wrap up the project and begin writing the final report. There are considerations when leaving the field. For one, the researcher may want to schedule additional meetings—focus groups or interviews—after the conclusion of the observation-based data gathering, in order to have the participants validate the findings or in order to follow up on interest areas not revealed through observation. The researcher should maintain the same high levels of trust that were so important during the initiation of the research project.

Recording observations using field notes. Field notes are a vital tool of the ethnographer. They are a written record of the researcher’s observations during his or her time in the field. Researchers would not be capable of remembering everything that they see during their observations, therefore keeping a notebook detailing the various activities of the community in which they are immersed is critical. Field notes are comprised of detailed records of observations as they occur. Over the course of compiling field notes, probative analysis also develops. Observers write down tentative and initial questions or conceptual possibilities alongside the field notes as they are collected. These are often referred to as “theoretical asides.” As observations and field notes grow in number, conceptual and theoretical elements may be combined and reflected upon in greater detail. More in-depth writing, focusing on explicit connections among and between the asides, are referred to as “observer commentaries” and represent a more sustained analytic treatment of the evidence collected. When the writer develops, in paragraph form, a rough draft explicating tentative themes, these writings are known as “in-process memos.”   

Strengths and limitations of ethnographic research. Ethnographic research has several strengths. First, ethnographic research observes the participants in their natural setting. In fact, to borrow terminology popular in quantitative social science, one would say that ethnographic research is ecologically valid. Rather than relying on a self-reported survey where someone reveals how they would act in a conflict situation in the workplace or relying on an experimental design where participants might role-play a workplace conflict in an artificial setting, ethnographers observe the participants in their natural environment. Second, ethnographic researcher stems from the thick, rich descriptions of the social actions of the group under investigation. Imagine the details collected by a researcher who is able to capture and describe group behavior as it unfolds and then follows up with informal interviews in order to gain the perspectives of the participants. This high level of detail is not something that survey data can replicate. Third, due to the extended observation periods in the natural setting and the cultural immersion that occurs with this method, ethnographers are able to gain deep understanding of the social activities of the group or cultural being observed.

Ethnographic research is not without its limitations. First, perhaps its most significant limitation stems from the fact that although one is able to develop an in-depth understanding of a group or culture, that understanding comes at the cost of generalizability. Just because a researcher can understand the social actions of one group or culture at a particular point in time does not mean that the knowledge derived from this understanding is relatable to other groups or cultures, which could then limit the applicability of the research. Second, sheer time and resource commitment required of a researcher to immerse herself and fully understand a culture is a limit. As noted previously, two years is not an exception but is, in many cases, normative. Finally, there exists the possibility that the researcher begins to over-identify with his or her research subjects. One example of this might involve a researcher deliberately leaving out accurate, but negative information about research subjects because the researcher does not want any aspect of the group to be seen in a negative light. Conversely, a researcher may deliberately exaggerate some characteristics to paint the group in a more desirable light. This over-identification is often referred to as going native, and it seriously jeopardizes the credibility of a study.  

Sample ethnography. If one wants to fully understand the process of ethnographic research, one should identify a research environment, initiate an ethnographic study, and begin data collection and analysis. To facilitate understanding of the ethnographic process, we describe Susan Weinstein’s (2007) study to clarify the process of ethnographic research. Weinstein (2007) spent three years gathering field observations, collecting written artifacts, and conducting informal interviews while studying how “nine low-income, African-American and Latino urban youths” wrote about gender and sexuality through “poetry, prose, and rap lyrics” (p. 28). In her own words, Weinstein states,

I wrote detailed field notes, working from notes jotted during the observations when note taking had been possible and unobtrusive. I conducted at least one sit-down interview with each writer, in addition to numerous informal conversations that were documented in field notes. I collected copious written and audiotaped compositions from all of the writers; in addition, I regularly read and took notes—and occasionally participated in—an online, interactive message board on rhythm-and-blues artist Alicia Keys’s Web site, where three of the writers traded freestyle rap lyrics and conversations about writing with peers from around the country and the world. All of the writers were fully aware that I was observing them and would be writing about them, all signed consent forms (or got parent’s signatures if the writer was younger than age 18 years), and I made a point of reviewing my analysis with participants as often as possible. (pp. 30-31)

In this single paragraph, Weinstein details many of the aspects vital to her study—the locations and settings where data was collected, the length of time spent conducting the research, information about the participants, consent gained, and efforts towards respondent validation. Respondent validation involves the researcher reviewing her tentative conclusions with the study participants in order to elicit their feedback regarding her interpretation of events. Weinstein’s research was guided by her belief that imaginative writing served as an outlet for the identity construction of urban youths. She found that gender and sexuality materialized in often contradictory ways in the writings of the youths and concluded those contradictory writings to be indicative of the complexity of the methods regarding these topics that youths receive from their social environment, such as family and friends and popular culture.

Ethnographic research is a challenging, yet rewarding, scholarly endeavor and a method to be used when one wants to develop as comprehensive and in-depth an understanding of a social environment as possible.

In-depth interviewing. In-depth interviewing is a qualitative method that fully situates the interviewee in the role of providing information to the interviewer. According to Lindlof and Taylor (2002), an interview is “an event in which one person (the interviewer) encourages others to freely articulate their interests and experiences” (p. 170). Ethnography may be a technique that is unfamiliar to many people, but interviewing is a process that most people have encountered at one time or another. A high degree of flexibility and variation characterizes the interview process. Qualitative interviewing can range from very structured formal interviewing to the loosely structured field interviewing that accompanies ethnographic research. This section begins with a discussion of the basic characteristics and goals of interviewing, followed by the steps for conducting an interview and the strengths and limitations of this method. We conclude with an exploration of a study relying on interviewing as the primary method.

Interview goals and characteristics. Qualitative research is inherently subjective. Qualitative research relies on its participants sharing their subjective understanding of certain experiences. Interviewing allows research participants to share their unique perspectives. Therefore, a primary goal of interviewing is for the research participant to answer the questions of the researcher. In essence, interviewing is the process of asking questions and receiving answers. Researchers want their interviewees to provide information about events or experiences that occur separate from the interview setting. Answers may take the form of stories or explanations. Interviews enable the researcher to understand the context and language forms of the social actors. Additionally, interviews allow the researcher to investigate events that he or she would otherwise not be able to access, such as closed meetings, past events, etc. The goal is to get the best possible data that will enable the researcher to successfully answer his or her research question or to provide an in-depth understanding of the social processes under investigation.

Regardless of the formality of the interview structure, an important characteristic of a sound interview is establishing a conversational tone during the interview process. Because trust is such a vital component of an interview, the interviewer should adopt a style that puts the interviewee at ease, and a conversational interview tone often goes a long way towards making the interviewee relaxed. According to Denscombe (2010), other important characteristics of the interview process include being cognizant of the feelings of the interviewee and the ability of the interviewer to tolerate silences. In day-to-day conversations silence is usually not tolerated very well, but when one is asking another to reflect on an issue, time is sometimes essential to allow the interviewee to think through an issue and to feel compelled to dig deeper in this analysis.

Time is another important element of the interview process. Interviews should last a reasonable amount of time. A researcher should not expect a participant to be willing to devote more than an hour of time for an interview, except in the most extreme of cases. Thirty minutes to one hour is considered a reasonable expectation for an interview. However, for some research questions, an even shorter interview may be enough.

Several ethical considerations confront a researcher during the interview process. As with all research involving human participants, gaining the consent of the participants is an obligation of the researcher. The researcher will also want to make sure that the interviewee understands how the researcher plans to use the information and how confidentiality will be ensured. It is common for the researcher to use pseudonyms for the interview subjects in order to mask their identities. A final ethical consideration is for the researcher to be aware of the potential for certain questions to lead the interviewee in a specific direction—questions posed in such a way that they elicit a particular answer from the participant. It is advantageous for the interviewer if the interviewee consents to an audio or video recording of the interview. Recording the interview can ensure the accuracy of the participant’s statements. With or without a recording, it is suggested that interviews be conducted in pairs—this way, one person can conduct and moderate the interview, while the other takes detailed notes regarding the interaction. Of course, the decision to use a second interviewer depends on the comfort level and consent of the participant.

Appropriate types of interview questions. Close-ended, yes or no, type questions are necessary from time to time during a qualitative interview, such as when gathering demographic information about your participants to write your methods section, but open-ended questions are recommended because they provide the interviewee with greater freedom in responding and offer the interviewer more data for analysis. Open-ended questions follow two common formats—non-directive questions and directive questions.

Directive questions are specific questions designed to discover specific responses. Examples of directive questions would be, “Tell me what kind of professor Tom Hall is” or “How does Tom Hall’s teaching style differ from April Chatham-Carpenter’s teaching style.” In both cases, the interviewer is asking the participant to address a specific point. Compare and contrast questions, as well as Devil’s Advocate questions, are examples of directive questions.

Consider the following non-directive question examples. “Tell me about a time when you experienced conflict in the workplace” or “Tell me a little about yourself.” In these examples, the interviewee is free to take the focus of the question in the direction of his or her choosing, and while that response will clearly contain specific elements, it begins with a more general approach than the directive questions.

Also, it is important to remember that not every question must specifically address the research purpose. For example, in order to increase comfort and gain the trust of a participant in the early stages of the interview, the researcher might consider asking a general question like, “Would you please tell me about yourself?” This allows the participant to grow comfortable sharing information with the researcher.

Finally, during the interview process, it is essential that the researcher is skilled at asking probing questions when more specific information is desired, asking clarifying questions when the information provided is unclear, or asking validating questions when the researcher wants to ensure that he or she interpreted the response of the participant correctly.

Steps of conducting an interview.   Five steps will help you through the process.

  1. Identify the purpose of your study and design the interview guide. The researcher needs to make sure that interviews are the appropriate methodology to employ in answering his or her research focus. If interviews are the appropriate data collection format, then the researcher needs to design a guide for how the interview will proceed. (The focus here is on research where interviews will serve as the primary data collection method and therefore are more formally structured than they might be during the impromptu interviewing that occurs during ethnographic research). A guide is just that, a guide to conducting the interview. It is a means for the researcher to organize his or her thoughts in order to ensure consistent approaches are taken across all interviews. However, it is just a guide and the researcher can add to it, take from it, or restructure questions as the need arises during an individual interview or over the course of multiple interviews.
  2. Identify the participants of the study and arrange times to conduct the interviews. If the researcher has selected a particular place or location to conduct his or her study, then the selection of the participants is likely a straightforward process. If, however, the researcher seeks to interview people who share common characteristics, rather than a physical location, selection of participants will likely rely on volunteer, snowball, or purposive sampling techniques. For example, suppose a researcher is interested in studying people who read comic books. In the process of identifying the participants, the researcher could gain access to a local comic book store and ask patrons if they are willing to participate in interviews, or the researcher could find one or two people who read comic books, interview them, and then ask them to help identify other potential participants (i.e., snowball sampling).
  3. Conduct the interview. Once participants have been identified and a suitable location (comfortable for the interviewee) has been arranged, the interview process can commence. Generally, a sound interview protocol would include introductions, research focus questions, and a conclusion.
    1. Introductions. Ask broad questions to increase the comfort level of the participants. In some cases a broad, introductory question might take the form of “Tell me about some of your communication strengths.” Questions of this type are known as biography questions and are designed to establish a conversational and comfortable tone for the remainder of the interview.
    2. Research focus questions. As the interview is underway, the researcher turns to those questions to which he or she specifically seeks answers. This is the main portion of the interview, and is where the researcher’s skills at asking probing and follow-up questions help him or her collect the desired information. As the interview moves from one topic to the next, it is helpful for the researcher to summarize the previous topic and solicit feedback from the interviewee before shifting to the next topic.
    3. Concluding the interview. It is common during this portion of the interview for the researcher to summarize main points in order to seek validation from the interviewee regarding the researcher’s interpretation of the responses. It is also common for the interviewer to ask the participant if he or she has any questions for the researcher. Also, always thank the participant for his or her time.
  4. Transcribe the interview. In order to accurately collect the data from the interviews, it is often necessary to transcribe the interviews. The transcription process can take as long, if not longer, than the interview process itself. It is essential that the researcher accurately portray the statements of the participants. Transcriptions also convert the data into a format that is often easier to analyze and interpret than the rough notes taken during the interview process.
  5. Analyze and interpret the data. The final step of the interview process involves analyzing and interpreting the data. Once the interviews have been transcribed, the researcher sorts back through the data in order to identify themes and common elements among the responses. Obviously, the researcher needs to keep the original research purpose firmly in mind when interpreting the data. The practice of analysis and interpretation is similar across the various qualitative methods, so a more lengthy discussion of analysis and interpretation will be presented in the final segment of this chapter.

Strengths and limitations of interviewing. Like so many other qualitative methods, one of the strengths of interview data is, quite simply, the depth of the information gathered by the researcher. Over the course of the interview, the researcher is able to probe, refocus, and follow-up on the various responses from the participant. Because of these characteristics, rich, detailed information is the product of qualitative interviewing. Another advantage stems from the flexibility of the interview process, and this flexibility also contributes to the depth of the information. Although you will likely follow an interview guide, the process itself is not as concrete as survey research questions. The researcher can adjust to the responses during the interview and guide the interview in the direction necessary to speak to the overall research agenda. It is also possible to augment and excise questions following interviews. For example, if the first three interviewees all talk about a specific event, it alerts the researcher to the importance of this event. In preparation for the subsequent interviews, the researcher can include a question to make sure that the previously mentioned event is addressed in future interviews. Interviews also allow for the acquisition of the participants’ subjective interpretation of the events being discussed. This is data in the actual words of the participants. The researcher does not rely on previously established categories but rather on the words and experiences of the interviewees. Finally, in many cases, interviews are the only means of discovering information about events that have already taken place. Interviews allow the researcher to uncover information that otherwise would not be available to him or her.

As far as the limitations of qualitative interviewing go, there are several that are worth mentioning. The sheer amount of time involved in setting up, conducting, transcribing, and analyzing interviews is daunting for many researchers. A one-hour interview may take three times that long to transcribe, particularly without the aid of electronic transcribing devices. In many cases, the interviewee may wander off-course during the interview process. The researcher has to balance the comfort level of the participant with the researcher’s need to gather relevant data. In some cases, brief meandering may be necessary in order to maintain a comfortable conversational flow between the interviewer and interviewee. There are, of course, other concerns. It is one thing to focus the interviewee on answering a specific question, it is quite another to deliberately lead the interviewee to a specific response desired by the researcher. Researchers must be fully aware of the impact and influence that they have on the entire interview process. As is always the case with qualitative data, the data collection instrument is the researcher; therefore, the limits of the researcher will affect the entire process.

Interview study example.

Cohen, M., & Avanzino, S. (2010). We are people first: Framing organizational assimilation 
experiences of the physically disabled using co-cultural theory. Communication Studies61(3), 272-303.

Cohen and Avanzino examined “how organizational members with disabilities experience and manage organizational assimilation in the workplace” (p. 272). They conducted interviews with 24 individuals with physical disabilities. The researchers employed snowball and purposive sampling to identify participants and “sixteen interviews were conducted face-to-face and eight took place over the telephone due to distance and time constraints” (pp. 280-281). From the twenty-four interviews, 140 pages of transcriptions were produced. From the data, Cohen and Avanzino uncovered eight concepts and two themes, and identified aspects of the difficult process of workplace assimilation, as well as various techniques employed by the study participants to successfully negotiate workplace assimilation.

All of the elements of qualitative interviewing are present here: a general research question focused on understanding rather than prediction and control; a small, purposely selected group of participants; and an in-depth analysis of the participants’ responses to develop themes and facilitate understanding of the assimilation process.

Research methods are not mutually exclusive, and although this study primarily used interviewing, the authors note that they also engaged in observer-participant activities. In fact, researchers often triangulate their methods, combining the next two methods to be discussed—focus groups and narrative interviewing—in conjunction with ethnographic research or qualitative interviewing to strengthen the quality of the study.

Focus group interviewing. The fundamental difference between interviewing and focus group interviewing is that focus group interviewing is designed to allow multiple participants to interact with one another. Regular interviewing often occurs one-on-one; focus groups often bring together 6-12 participants in order to gather data as they interact with one another. Many people are familiar with marketing or political focus groups designed to uncover people’s attitudes towards a particular product, political figure, or idea, but fewer people are familiar with scholarly focus groups. Research focus groups may be similar in number of participants and duration of the interaction (90-120 minutes) to these others types of focus groups, but their purpose is not to gauge attitudes about a brand or political figure. Much like a regular interview, a focus group interview is designed to elicit information from the participants but is arranged in such a way that the participants are able to openly engage in discussion with the other participants. This experience, while often difficult to moderate, can provide a wealth of data. This section includes a discussion of focus group characteristics, moderator concerns, steps in focus group interviewing, and strengths and limitations. It concludes with a look at research employing focus group methodology.

Characteristics of focus groups. In addition to the number of participants and the length of time required to conduct a focus group, other important characteristics distinguish focus groups. Focus groups allow the researcher to interview several people at once in a format that resembles a purposeful discussion. Focus groups allow researchers to gather information from a group of people in a single setting. Some of the characteristics shared with regular interviews include the designing of an interview guide and involving two researchers for the process (one to moderate and another to take notes). In the case of the interview guide, it should be clearly developed but will likely not be as lengthy as the guide for a one-on-one interview. Because focus groups allow the participants to interact with one another, a few questions by the moderator may be all the prompting needed to elicit discussion. Due to the collaborative nature of focus groups, the moderator may only need to initiate the discussion and then can spend the majority of his or her time managing and focusing the ensuing discussion rather than constantly interjecting new questions. In some cases, the interactions among the group may emphasize points of agreement, as several of the participants add on to a topic, idea, or event that has been introduced. On other occasions, the group interaction may result in points of contention, enabling the researcher to see where participants have very differing perspectives on the research questions and topic.  

Moderator concerns. The moderator’s job is much more difficult in a focus group than it is in a one-on-one interview. A focus group moderator has to manage multiple personalities rather than a single personality. It is still important to establish a conversational tone among the participants, but it is also necessary to pay close attention to the various personalities of the group. Is one person dominating the discussion? Are two people ganging up on another member? Are tensions running high among the group? Is one person overly shy and unwilling to open up? These are just a few of the concerns, characteristics, and mannerisms that a focus group moderator may need to address over the course of the focus group interview. Practice is the best tool for learning when to allow disagreements and when to cool discussion down before arguments develop. Remember, the express disagreement that stems from constructive conflict is very different from the hostility associated with destructive group conflict. The focus group moderator must also insure that the group remains focused and does not wander too far from the original intent of the moderator’s topic or question. The moderator should also refrain from displaying any bias over the course of the focus group.

Steps of conducting a focus group. Some of the standard steps of designing and conducting a focus group are as follows:

  1. Identify the purpose of your study and confirm that focus groups will be the most useful method for your study. Researchers will want to consider their overall research focus and then construct a list of questions that will provide the best opportunity to elicit relevant responses from the participants. You should have a clear purpose to the focus group questions, but you also need to have the flexibility to adapt as the situation merits.
  2. Recruit participants. Once the researcher has decided which participant attributes are essential to the study, he or she needs to initiate the process of recruiting participants. Once again, snowball sampling and purposive sampling are often the most useful methods of recruitment. Even though a desired number for a focus group is between 6-12, the researcher should select a number that he or she feels capable of moderating. Also, just because people say that they will show up does not mean that they will. It is better to have too many people show up for your focus group session, and have to turn a few people away, than to have too few people show up. It is acceptable to over-recruit by a person or two.
  3. Conduct the focus group.
    1. Introduce purpose of the group, participants, and explain the expectations and ground rules for the discussion. The first segment of the focus group should begin with the researcher/moderator explaining the goals and purpose of the focus group, followed by a presentation of the ground rules and discussion expectations. This is a good time to express the desire for respectful conversation and equal sharing of information. It is also a good time to allow the participants to introduce each other if they are not familiar with one another.
    2. Ask questions and moderate the ensuing discussion. Questions should be open-ended and may be either directive or non-directive depending on the needs of the researcher. It is during this segment that most of the moderator’s skills will be put to the test as he or she strives to keep the group on task, sharing equally, and clarifying points of contention or agreement.
    3. Conclude the focus group. Similar to one-on-one interviewing, the moderator should allow time for the participants to clarify or elaborate on any of their previous statements. Participants should also have the opportunity to ask questions of the moderator. Finally, make sure to thank the participants for taking the time to be a part of the focus group.
  4. Transcribe the focus group data. One of the challenges of focus group research is clearly differentiating the various participants, who in some cases will talk over or interrupt one another. This is one of the reasons that it is good to have a moderator, a note taker, and multiple recording devices (to catch all the voices) throughout the focus group. Transcribing data of this sort can be a time consuming process. When it comes time to analyze and interpret the data, detailed and accurate transcriptions are a necessity.
  5. Analyze and interpret the data. Once the notes have been transcribed, the researcher sorts through the data in order to identify themes and common elements among the responses. Obviously, the researcher needs to keep the original research purpose firmly in mind when interpreting the data. The practice of analysis and interpretation is similar across the various qualitative methods, so a lengthy discussion of analysis and interpretation will be presented in the final segment of this chapter.

Strengths and limitations of research focus groups. Obviously, the greatest strength of focus group methodology is the interplay among the various focus group participants. The healthy give and take among the participants serves as a fruitful generator of data. In fact, the primary reason for selecting focus groups over one-on-one interviews is so that the researcher can record several people interacting regarding the same topic. Provided that all of the subjects of the focus group participate, the researcher can gather a multitude of opinions and ideas on similar topics. Another advantage is that you can collect a relatively large amount of data in a brief period of time. In the time that it might take to conduct two individual interviews, the researcher can conduct a focus group with 10-12 participants. Although the depth of the data as compared to individual interviews may suffer, the breadth of the data and the discussion of common topics are extremely beneficial.

There are limitations to focus groups, several of which have already been highlighted. A skilled moderator is a necessity or the group can quickly get off task and produce information that does not relate or address the original purpose of the research. There also is the potential that overly dominant group members will lead the discussion in ways that might not represent the feelings of the remainder of the group. Finally, group members might either go along to get along or deliberately disagree (playing Devil’s Advocate)—neither of which will lead to the most accurate data. According to Morgan (1997), there is always the possibility that the moderator has overly influenced the groups. Focus groups do not take place in the natural environment; they are artificially constructed. This, in turn, impacts the ecological validity of the research.

Focus group example.

Hundley, H. L., & Shyles, L. (2010). US teenagers’ perceptions and awareness of digital 
technology: A focus group approach. New Media & Society12(3), 417-433.

Hundley and Shyles conducted focus groups with 80 middle and high school teenagers. “The chief objective of this research was to further our understanding of what young people think about digital devices and the functions they serve in their lives” (p. 417). A total of eleven focus groups were conducted with five to nine students in each group. Hundley and Shyles utilized both formally structured and semi-structured focus group interview protocols. The formally structured segment consisted of specifically prepared questions, while the semi-structured portions “allowed students to speak freely, elaborate, ask questions and join in group discussions” (p. 419). According to the authors:

Some focus group questions aimed at understanding what digital technology the participants possessed, used, or wanted to own. Others asked specifically about various devices such as game systems, computers, cameras and cellular telephones, to develop an understanding of participants’ familiarity with each piece of technology. Questions also inquired about how teenagers spent their time with various digital media devices and what they liked to do with them. . . Typically, the conversations were dominated by talk about the internet, specifically social networking sites. (p. 420)

From the focus groups, Hundley and Shyles were able to identify several common themes among the 80 participants. Themes included high level of awareness regarding the various types of technology, a lack of awareness regarding amount of time actually spent using the technology (nearly all underestimated time spent), awareness that digital devices help them socialize, and the risk of having personal information available online. Hundley and Shyles found that their research was consistent with the existing literature regarding teens and technology.

Autoethnography is a relatively new qualitative research method that is generating a great deal of interest. It is based in ethnography, meaning that it, too, is an attempt to understand and describe the insiders’ cultural perspectives—i.e., how insiders construct their world view/culture. Like ethnography, it is also holistic and naturalistic, rather than trying to isolate what is studied and control it. Finally, like ethnography, it requires some degree of participant observation, but in this case, the observer may the reader, not just the researcher.

While there is no one definition of autoethnography, it is the study of some aspect of culture from the author’s personal experience and perspective. Examples of topics where authors have shared their personal experience through this method include surviving breast cancer, an eating disorder, depression, being of multi-ethnic identities,  the process of transitioning from woman to man as a transgender, and much more. The researcher is the subject of study, the key informant. The method is similar to narrative data collection. The researcher/author tells h/his story on a topic or issue the person feels warrants others to learn about from an insider view. The author role is reversed from being a researcher first and then an author to being a story teller first. If the story is not compelling, the research effort has failed.

Unlike other qualitative research that focuses on description, the goal in autoethnography is not just to describe but to evoke feeling and deeper understanding. This takes the view of knowledge as being an embodied experience, not just observation. To do so, the researcher must make h/himself vulnerable, sharing a great deal of self-disclosure and demonstrating a great deal of self-reflexivity.

Nevertheless, autoethnographers still value systematic methods used in other research methods:

I start with my personal life. I pay attention to my physical feelings, thoughts, and emotions, I use what I call systematic sociological introspection and emotional recall to try to understand an experience I’ve lived through. Then I write my experience as a story, By exploring a particular life, I hope to understand a way of life… (Carol Ellis & Arthur Bochner, 2000, p. 737)

If you imagine research methods as on a continuum, quantitative laboratory research would be on one end of the spectrum, and autoethnography would be at the other end of the spectrum, followed by performance studies and other more artistic ways of knowing. Because this method requires a unique set of writing skills, we do not cover it in full here, but offer websites for those who might like to learn more.

For an overview: http://www.qualitative-research.net/index.php/fqs/article/view/1589/3095

For a focus on analytic, rather than evocative autoethnography: 
http://web.media.mit.edu/~kbrennan/mas790/02/Anderson,%20Analytic%20autoethnography.pdf 
For an example of autoethnographic research as researcher ethics: http://jrp.icaap.org/index.php/jrp/article/view/213/183

To perform autoethnography: http://eppl604.wmwikis.net/file/view/spry.pdf

Media portrayal of what autoethnography is: 
http://www.youtube.com/watch?v=pb50nPHgI04

Data Analysis: Interpreting Results

Collecting information is only the first of two parts in the research process. In qualitative research how one interprets the results is particularly salient. Recall that the world-view or epistemological approach of qualitative research is the assumption that multiple meanings are always possible and present and that meaning is created, it is perceptual, and influenced by context. It is not just observed, nor is it considered universal or objective. Thus, in the second part of the qualitative research process the researcher purposefully and explicitly makes meaning of the study by applying methods to reveal patterns in the data. As discussed in chapter one, making meaning is what people do when they interact every day. They negotiate and construct perspectives through the exchange of verbal and nonverbal cues. In research, the author makes meaning through a negotiation with the verbal and nonverbal messages or data collected. The researcher’s charge is to make every effort to make fair and insightful sense of what might currently be a large pile of data.

As this description suggests, because there is room for multiple interpretations, the researcher has a responsibility to analyze the data or information gathered in a highly systematic fashion, drawing from previous methodologists’ recommendations and guidelines. To be systematic means the researcher cannot simply focus on data that is the most convenient, or anecdotal and examine it in a half-hazard fashion. Ideally, the researcher must make every attempt to consider all the information even though it would be impossible and not useful for the researcher to include all the information in the final report of the study.

Common Steps in Qualitative Data Analysis

Analysis methods for qualitative research have several commonalities. Recall that in all qualitative research, the data is words and behaviors, not numbers. It may be in the form of artifacts such as newspaper clippings or diaries from the field of study, field notes you have taken, transcriptions of one-to-one interviews, focus groups, and/or participant’s more naturally occurring conversations. In the analysis phase, the researcher’s job is to select analysis tools  that seem to be a good fit for the type of data being examined and the objectives of the study and then to review all the data in a consistent fashion. From this, the researcher seeks to synthesize the material by organizing it into categories and then identifying connections among the many categories that will make visible a more narrow, manageable focus to make sense of the information. 

There is a variety of methods to conduct qualitative data analysis. Some are more complex, others more accessible, but it is helpful to remember that at their essence, the meaning making process for all of them is to look for patterns -- themes and patterns among the themes by “comparing and contrasting parts of the data” in a series of steps (Keyton, 2011, p. 62). Thus, before reviewing specific types of qualitative methods, we are able to identify basic steps common to all of the methods.

According to qualitative communication scholars, Thomas Lindlof and Bryan Taylor (2002), the overall analysis process involves three basic tasks: “data management, data reduction and conceptual development” (p. 211). In data management, tools to code and categorize data are used to help the researcher gain some control or order over what could be an endless volume of information. From there it becomes easier for the researcher to see that some data may be more central to understanding the results than other data. Data reduction is then when the researcher begins to assign prioritized value to the data, reducing the size of the focus of analysis for a closer look. This does not mean any information is to be thrown away. Remember qualitative researchers believe meaning is created in context. Although it is necessary practically to try to distinguish what is most salient to focus on, the other information may in time provide a rich context for more complex, subtle interpretations. The researcher will want to later return to this secondary data for such considerations. Conceptual development should emerge from accomplishing the prior two tasks. This is where and how the concepts and themes -- or meanings in the study emerge. They are grounded in the data from the specific social context of study as well as influenced by the researcher’s review of previous research and qualitative theory. Together the tasks of data collection, data management, data reduction and conceptual development emulate the inductive nature of qualitative research. The researcher studies communication in a smaller, more specific context, gathers extensive data, organizes it and reduces it even further to smaller kernels of knowledge, and then returns the kernels to the larger social and theoretical contexts for further thought about the significance of these interactions in everyday life.    

In applying Lindlof & Taylor’s three steps for qualitative analysis, we break down the process further, adding two additional steps to make the process more explicit. We add an introductory step called data immersion and a concluding step called evaluating results. While we present them as if they are discrete steps, the reader should know researchers begin to think about these steps before all the data is collected and may return to the data collection phase after conducting preliminary analyses. The reader should also be forewarned if you read other descriptions of this process the steps may be divided a bit differently or labeled differently. We tried to do what was most streamlined. In the end, the various descriptions are really about the same basic process.

Member checks add to the credibility of a study. They better assure the participants’ perspectives are respected and that the interpretations are grounded in their perspectives (Lincoln & Guba, 1985). Useful results from member checks are not just results that say, “yes, this is good,” or “you made a mistake in the spelling of my name,” but rather feedback that helps the researcher see the findings through the participants’ eyes, thus extending the depth and breadth of interpretations and resulting theory proposed. Another value added from this method is that it gives some ownership of the results back to the participants. They are empowered to edit and add their input (Lincoln & Guba, 1985).

For example, in a study of African-American women’s self-esteem, DeFrancisco and Chatham-Carpenter (2000) conducted member checks in recognition of their position as White women interpreting the stories of women of color. The meetings caused the researchers to reframe one of the primary themes taken from the study to focusing on the effects of racism on the women’s self-esteem to focusing on the demand for respect. The shift in words from the researcher’s (racism) to the participants’ (respect) may have seemed small, but the former framed the participants as powerless victims of a racist society, the latter framed them as strong women demanding fair treatment.

  1. Data Immersion: Get to know your entire data set closely. This includes the participants’ contributions as well as your field notes during data collection. A close-read, includes reading and rereading the data set multiple times, line-by-line, perhaps in different orders (Lindlof & Taylor, 2002). There is nothing that can be substituted for building a close awareness of what you have gathered. You will likely gain a new insight every time you review the material. You need a view of the whole before you can begin to sort the data. As you do so, look for holes – is there anything you need to go back to the field of study to gather? Use a concept we discuss in more depth later called reflexivity (Ellis, 2004). Critically monitor and write down your observations, reactions and feelings as you process the information collected, they may influence the directions you take next. The instructions for taking good field notes described under data collection above will be helpful here, as well. It is a good idea to make a complete copy of your data and archive it in case the copy you work from becomes damaged or lost.
  2. Data Management: This step is vital to making meaning of the data. And it is one you will do over and over, each time tweaking it a bit better to fit what you think the participants were saying or doing. You the researcher, must organize your data into categories or themes so that you can get a handle on the data -- work with it more effectively and begin to see what it means. According to communication studies researcher Joann Keyton, “a theme is a conceptualization of an interaction, a relationship or an event” (2011, p. 313). Data management consists of basically two steps: a) identifying and defining what counts as a piece of data. This is called the unit of analysis; and b) coding – creating labels to categorize or place the individual pieces of data under. Categorizing is how you decide to group your data into subparts for closer analysis. Qualitative scholar Susan Spiggle defines the categorization process as, “identifying a chunk or unit of data (e.g., a passage of text of any length) as belonging to, representing, or being an example of some more general phenomenon” (1994, p. 493). It is putting together items that seem similar to each other.
    1. The chunk is called a unit of analysis or unit of data. Identifying one’s unit of analysis is necessary in qualitative as well as quantitative data. It helps to define what level or size of data the researcher is focusing on – to identify the boundaries of individual units of data being considered. The unit of analysis is the smallest level of the data from which the researcher sorts the data and thus making meaning. The goal is to have the same or similar size for each chunk of data. The size depends on the goals of the study. For example, if a researcher is studying how people use conversation to create and maintain their identities, the researcher’s unit of analysis is typically a person’s turn taken in a conversation. But in most theme/categorization analysis of qualitative data, the unit of analysis can be more flexible. It can be a word, a phrase, a complete sentence, a conceptual theme, a nonverbal expression, a communication episode or event such as episodes from reality television, full texts, such as speeches, and more. Since the size may vary, you may find recommendations from researchers Yvonna Lincoln and Egon Guba (1985) helpful for identifying your unit of analysis. They suggest the unit have two criteria: the first is noted above – that it be the smallest piece of recognizable information, meaning that it does not require other contextual information to be understood. Second, the piece of information should be heuristic, meaning it should help with understanding. It should help address the research question or objective of study.

      For example, imagine you are studying how young adults perform their identity by what they choose to post on their Facebook or another social media site. Your unit of analysis will likely be each verbal or nonverbal indicator about the participants’ identity. So, if a person posted, “I’m a 21 year old woman who loves to laugh, but I am dead serious about the type of man I want in my life.” The following is how the sentence might be broken into individual units, or pieces of information about her identity: “I’m a 21 year old/ woman/ who loves to laugh,/ but I am dead serious/ about the type of man I want in my life/.” Thus, from just one sentence, the research can glean 5 units of data representing 5 different categories or themes: age, gender, personality, goal (looking for a mate), and sexual orientation (heterosexual).  

    2. Once the unit of analysis is determined, the researcher can chunk up all the data into units and begin coding the data – organizing it into groups under labels. A coding scheme, is a sort of shorthand device to label and separate the data (Landlof & Taylor, 2002). When finished, the codes will help to reveal categories of concepts or themes and codes make it possible to retrieve the data more easily for further analysis.

      So, for the example of social identity in Facebook above, as you review the data collected from other participants’ sites and your interviews with them, imagine you notice a tendency in the descriptors that seem to refer to the participant’s demographics: gender/sex, sexual orientation, age, etc. These demographics may each become a category where you will compile the individual instances that demonstrate that category. A post from a person who identifies as male describing how he lifts weights and strives to attain muscle mass might be categorized under gender, and specifically masculinity. A person who posts a message about being a women who competes in a race for wheel-chair users might be categorized under physical abilities, etc.

      The code or coding scheme is to look for demographics, the resulting categories or themes are gender/sex, sexual orientation, age, etc. In the study of coming-out narratives, the code is degree of agency, the categories are the specific degrees noted in the data. In the study of relational conversation, questions are the code, the specific functions of questions are the categories. Coding helps to identify whole families of categories and sub-categories. They help the researcher not only sort, but display the data in visual ways to make the meaning more apparent.

      We offer two cautions regarding oversimplifying this process. First, while you may see preliminary codes and categories emerge from the data, the final ones may not be the same. It is important to keep one’s mind open to considering alternative or extended coding systems otherwise one may only see what s/he wants to see. While there are multiple ways to organize data and endless sizes or levels of latter categorization possible, there will likely be a way that makes sense to you. There is also no set number of categories required. You should have as many categories as it takes to account for all the data possible. In the end, you should feel the categories do a good job of representing the data as a whole, both in terms of its similarities and in terms of its differences (Lincoln & Guba, 1985).

      Second, a warning about labels. Qualitative researcher use the terms coding schemes and categorizing to refer to the process of organizing data in qualitative analysis, but they do not always use the two terms in the same way. Some researchers use the terms interchangeably. Others claim there are important distinctions between the two terms. We follow Lindlof and Taylor (2002) who argue researchers first code the data into groups and then look for categories (how the groups connect in terms of concepts, themes, constructs). Codes are descriptive of the data groupings, whereas categories and themes are more interpretive – they tell what overall meaning the researcher makes of the groupings. Themes and category systems are the result of coding (Saldana, 2009). For example, a code may be chunking the data by participant demographics, which results in a categorization system of data that emerged according to race/ethnicity, social class, age, and gender/sex. A theme for the category of race/ethnicity might be “participants identify with homogeneous others” or “participants celebrate multicultural identities.” Each are conclusions or themes drawn from data within the category of race/ethnicity.

      To further assist you in identifying codes and categories, two other researchers proposed a list of questions you might find useful (Lofland & Lofland, 1995).

      Tips for coding and categorizing data:

      • Ask yourself a set of questions for each piece of data:
        • What is this? What does it represent?
        • What is this an example of?
        • What do I see going on here? What are people doing? What is happening? What kind of events are at issue here?
      • Other questions to help determine how to organize the data:
        • If something exists, what are its types?
        • How often does something occur?
        • How big, strong, or intense is something?
        • Is there a process, a cycle, or phases to the topic of study?
        • Does one thing influence another?
        • Do people use the interaction in specific ways?

      (See also Sociologist Graham Gibbs’ lecture, “What is coding for?” http://www.youtube.com/watch?v=5xM-9yuBhMc)

  3. Data Reduction: If the data set is relatively small or the initial categories seem particularly salient, the researcher may not need to proceed to this step of refining the focus of analysis. However when conducting extensive ethnographic research it is particularly necessary to reduce the focus of analysis to a size and framework the researcher can better manage. Framing or frame analysis is selecting a focus or lens from which to analyze and present the data. When doing so, the researcher acknowledges that multiple frames are possible and that one’s unique perspective may influence how one positions and interprets data. One has to make choices and this fact should be documented and rational provided for the choices made. For example, one’s roles in society may influence the frame taken in analysis: a student, a psychology major, a sociology major, a communication scholar, a parent, a single person, etc. (Grbich, 2007). The goal is to develop in-depth quality in the final analysis, not quantity. Thus the researcher may begin to identify primary categories or themes and secondary ones that may or may not end up in the final report.

    If we return to the example of studying participants’ identity construction on social media, the researcher may decide reporting people tend to describe themselves according to demographics is not very insightful or useful. Perhaps the comments suggest a particularly interesting connection among the participants’ comments about their gender, sex, and sexual orientation, and that given the researcher’s review of previous research and/or training, the researcher decides this focus would make a more useful contribution to the state of knowledge. The researcher is then framing the study around themes or categories tied to gender, sex and sexual orientation. Other demographic information such as race, ethnicity and age may be related in the participant comments and serve as a secondary level of themes or categories to be explored.

    Throughout the analysis process, but certainly while conducting data reduction, the researcher should be looking for exemplars – examples such as quotes from the data that vividly illustrate the themes and categories proposed. The examples should emulate the boundaries or characteristics the researcher has used to distinguish the themes or categories. They should make the themes or categories come alive for the reader (Ryles, as summarized by Geertz, 1973). These are what will be presented in writing the report to represent the qualitative data reviewed.

  4. Conceptual Development: The analysis process would not be very meaningful if the researcher stopped after creating a list of categories. The researcher now needs to determine what meaning to make of the list. In this step the researcher attempts to integrate the categories to consider what they mean when examined together. The researcher looks at how they might relate to or influence each other (Lindlof & Taylor, 2002). This phase is a meta-analysis – an examination of the examination of data – the creation of codes, categories and themes created in the previous steps. In grounded theory, described later, this is called axial coding, but the process and goal is the same across methods – attempting to connect the codes, categories and/or themes.

     

    The process used is basically to repeat the same steps but on a higher level of analysis: review all the codes, categories or themes (instead of the individual instances of each) in comparison to each other and attempt to understand how they relate to each other. It is as if the researcher is constructing a framework or umbrella in which to locate the individual codes, categories and themes. There are many ways researchers may attempt to do so. The manual methods suggested previously for color coding, creating grids or other visual depictions of the categories and themes can help make the connections visible. Researchers may see that the categories or themes fit well with an existing theory and so place them within that theoretical context, they may see that together the parts suggest a new theory or conceptualization of the issue being studied, they may see a metaphor emerge that helps to explain how the parts fit together, etc.

  5. Evaluating Results: The qualitative researcher is not done with data analysis until appropriate tests are applied to increase her/his confidence in the interpretations rendered. While we list this step last for clarity and emphasis, researchers should adopt the practices described here throughout the study. This is important to raise the rigor, credibility and ethical integrity of a qualitative study given that multiple interpretations of data are possible. Otherwise, how do the researcher and readers know to trust the interpretation presented? There are a number of tests or methods a researcher can use. We point out five commonly used methods, some of which have already been introduced.

    • Triangulation: As defined in the first part of this chapter, triangulation is approaching the study from multiple perspectives to enhance the rigor or integrity of the results of the study. It can include using multiple methods of data collection, gathering multiple sources of data, and for analysis it can include having multiple researchers or coders for the data, and/or drawing from multiple academic disciplines to frame the study. The idea is that if shared meanings emerge from multiple directions of data collection and analysis, those meanings are likely more sound. Sociologist Norman Denzin (2006) proposed this method to directly refute the criticism that because multiple interpretations are possible in qualitative research, the findings proposed from a study are unreliable or invalid. 
    • Representativeness: The researcher should be sure there are multiple exemplars – specific examples of quotes or behaviors that support each theme or conclusion formed. If the researcher cannot come up with enough strong examples, the claim is likely not strong enough to be warranted.
    • Member check: After preliminary or final analyses, the researcher shares the interpretations with participant members of the study to solicit feedback. The sharing can be in the form of face-to-face interviews, focus groups, or written data summaries. The goal is for the researcher to find out if the interpretations rendered ring true to the participants. Do they feel the results reflect their lived experiences? The researcher may want to use a detailed standard list of feedback questions or ask for a more holistic reaction to the research summary. After obtaining input, the researcher implements the information gathered to rethink the findings and/or cite as support for the claims in the study.
    • Transparency: Transparency in research is the expectation that researchers will make the entire process of their work explicit, openly sharing the process of meaning construction with the readers (Hiles, 2008; Seale, Gobo, Gurbrium & Silverman, 2004). This criterion emerged out of a need, as previous qualitative and quantitative research was often not transparent, in part due to page limits and related costs for academic journal publications. Thus research conclusions and what academia comes to call knowledge appeared as if from a vacuum, free of individual decision-making and perceptual influences that are always present in human endeavor, and prevented others from testing out the results further. Transparency is an ethical consideration. When researchers clearly document the steps taken in data collection and analysis and share these with the reader, s/he should be able to better understand and visualize how the conclusions were formed. If this clarity is not present, the value of the study and the researcher’s integrity may come into question. 
    • Reflexivity: An effort to examine how one’s own thoughts, feelings, and behaviors might intermingle with phases of the research process (Bochner & Ellis, 1992). Reflexivity is the recognition that the researcher is a part of what is being studied. The researcher’s unique cultural lens will necessarily affect the research process. Taking the time to place oneself and one’s values and possible biases under examination better assure this inevitable influence is not abused and that not only the researcher, but the participants know the nature of the researcher’s likely influence on the study. Careful field notes and keeping a diary or log during the analysis process will help the researcher examine this criterion.
    • Grounded Theory – Grounded Theory is actually one of the individual analysis methods described below. Its mention here speaks to the pervasive influence of grounded theory on the general process of all qualitative research. The key point regarding using grounded theory as a criterion to assess rigor and integrity is that the researcher should refrain from using her/his words to label themes/categories, and rely wherever possible on the words and images conveyed by the participants. The example above on studying self-esteem from African American women’s perspective illustrates why this is so important. The two White women researchers had initially inadvertently assigned their own label for a theme as racism rather than “respect” which the participants used.

These five steps represent the general process of conducting qualitative analyses. What follows is information about the specific analysis methods most commonly used: thematic analysis, grounded theory, and content analysis. We will also briefly introduce the reader to two methods that focus more specifically on how the verbal and nonverbal messages under study were constructed – discourse analysis and conversation analysis.

Thematic Analysis. This is the most general, easily accessible method of data coding or categorizing. The reason this method is more accessible is that the rules for the general process of analyzing data as described above are more relaxed. This does not mean thematic analysis is unsystematic however. The researcher still defines the unit of analysis and data is coded first to organize it. From the categories the researcher looks for a theme within each category and then an umbrella theme or themes that might connect the individual themes. The concluding themes should best reflect the data as a whole.

In a comparison of grounded theory and thematic analysis, Mohammed Ibrahim Alhojailan (2012) concluded thematic analysis is a comprehensive method, just as is grounded theory, however “It provides flexibility for approaching research patterns in two ways, i.e. inductive and deductive” (p. 39). In its purist sense, grounded theory requires data to be inductive only. In thematic analysis, the themes may be based on information gathered from interviews and previous research and theory. And, the data does not have to be collected at one time. “This makes the process of thematic analysis more appropriate for analyzing the data when the researcher’s aim is to extract information to determine the relationship between variables and to compare different sets of evidence that pertain to different situations in the same study” (p. 39). 

Characteristics of Thematic Analysis: Theme analysis is often used to study texts, both written and transcribed oral texts such as interviews or focus group discussions. The themes can come from the research questions and objectives guiding the study, from previous research or theory presented in the literature review, from the researcher’s standpoint as a certain gender and sex or ethnicity, social role, etc., from the data itself gathered in the study, or what is most common is from a combination of pre-existing influences and meanings that emerge from the data.  “This makes the process of thematic analysis more appropriate for analyzing the data when the research’s aim is to extract information to determine the relationship between variables and to compare different sets of evidence that pertain to different situations in [the] same study”   Alhojailan (2012, p. 39).  

Steps of Conducting a Thematic Analysis: While the overall steps are as described above for qualitative data analysis, in thematic analysis steps 3, 4 and 5 take on a particular process. The researcher must answer the question: When do comments or behaviors become a theme? Literally anything could be claimed as a theme, so what makes a given researcher’s claims acceptable? While the comment or behavior needs to occur repeatedly in the data, counting repetition alone is not enough. Interpersonal Communication scholar William Owen (1984) suggests the researcher will know s/he has a theme when it meets the following criteria: recurrence, repetition, and forcefulness.

  • Recurrence means that at least two parts of the data have the same meaning, although the meaning may be expressed in different words. The researcher looks for a pattern in the relational or underlying meaning of a message.
  • Repetition is about frequency – that the same key words, phrases or sentences are mentioned again.
  • Forcefulness refers to the degree of emphasis conveyed in the message. Does the way an idea is said or written suggest it is important to the speaker? This can be through vocal inflection, timing, volume, and/or emphasis placed on words or phrases in written or oral form.

While Owen would require that the data meet all three of these criteria, it may not always be possible to do so. What is important is that the researcher show evidence in the text to support her/his claims, and that they speak to the underlying meanings being expressed. There will always be other meanings in a message; the researcher’s job is to best assure the themes selected seem primary, rather than secondary. 

Owen offers an example from an analysis of a daily log a female college student was asked to make about her relationships. This one is about a high school friend. The short passage reveals all three criteria simultaneously:

Day One: She is an ideal friend. I haven’t really known her for very long, but it seems like we jumped into the middle of a relationship. I feel like I’ve known her forever. 
Day Three: That night we burnt chocolate-chip cookies and drank white wine, and for the first time since Bud died (her brother), I had someone I could relate to.
Day Five: Special is the word-of-the-day.

The thematic concept Owen claims is relational uniqueness. The references to “ideal friend,” jumped into the middle of a relationship,” “feel I’ve known her forever,” I had someone I could relate to,” “special is the word-of-the-day,” “It’s like Debbie and I share a secret,” “she is so special, we are so special!” “We have the perfect relationship,” all suggest the theme of uniqueness. Recurrence is apparent as is repetitious, with the word “special” being used three times in one entry. Forcefulness is also apparent with the italicizes used as in “for the first time,” and “special.” Comparing the current relationship to the comfort of her relationship with her brother who has now died is quite powerful.  (For an online lecture on the steps of theme analysis, see sociologist Graham Gibbs, University of Huddersfield, UK).

Strengths and Limitations of Thematic Analysis. Because this method is a general and relatively simple (but not fast) process, it is the most widely used method for analyzing qualitative data. It is a good choice for novice qualitative researchers because the process is easy to understand. It mirrors the social cognition perceptual process we humans engage in every day: organizing countless types and sizes of data into categories to make sense of the world around us. Thematic analysis can be less time consuming than other methods for qualitative data analysis, although to be done well, it still requires repeated reviews of all the data and a willingness to sort and resort data according to appropriate themes. And, it can be applied to analyzing all sorts of data from looking for themes in artifacts, such as news stories about an event, to looking for themes across interviewees’ comments about a particular topic, to looking for themes in how individuals who share a cultural identity tend to exhibit similar communicative behaviors or categories. It also works when there is a large volume of data, which is not as easy to do with other qualitative methods. These strengths of the method also become limitations. Because it is so general, it is sometimes criticized for not being systematic enough. While not a failure of the method, some researchers using thematic analysis fail to fully account for their thinking that went into creating the themes claimed as results of the study. A limitation of the method itself is that the meanings of human behavior are interdependent, not independent of other being or the physical and social context. Thematic analysis calls for sorting data into discrete categories, whereby the same comment cannot be placed under two themes. Yet, as the example below will likely suggest to the reader, comments or behaviors often seem connected and could be placed in multiple categories. The themes are almost always interdependent, but the researcher may not acknowledge this. And so, research using thematic analysis is more easily open to criticism regarding the relevance of the themes identified and the bias or hidden agenda of the researcher’s choice in framing the data within the selected themes.       

Example of Thematic Analysis:

Pohl, Gayle, & DeFrancisco, Victoria. (2006). Teaching through crisis. The International Journal of Diversity in Organisations,[sic] Communities & Nations

In a study of why some college instructors addressed in their classrooms the events of the airplane bombings in the U.S. on 9/11 of 2001, their comments suggested the following themes: they felt they had no choice but to address the event; they needed to make sense of the events both for themselves and their students; and they felt the events fit their course content (Pohl & DeFrancisco, 2006). Below are sample comments that when sorted suggested the three themes.

  1. Instructors felt they had no choice but to address the event:
    • Because I felt I had to both for myself and for my students. The event was too large and had too many ramifications to ignore. I knew that it was going to be an issue that we would be dealing with for a long time.
    • I really wrestled with the decision. It seemed that it would be impossible, if not completely inhuman, to try to ignore.
    • Both the students and I seemed to need to talk about it and explore the events surrounding 9/11.
    • It was a historic, painful, and socially critical event.
    • A lot of people, including students, were experiencing a variety of strong emotions. I think it’s important to open discussion for those who want to talk and those who might benefit from listening to others’ ideas. Our students look to us for guidance – we should provide it.
  2. Instructors needed to make sense of the events both for themselves and their students:
    • I teach in the state of New York. During the days and weeks that followed 9/11, a few members of my classes were mobilized as part of the National Guard. Other members were “absent” mentally or physically because they had found out, or hadn’t found out, about whether their family members and dear friends were alive. These issues were so in our faces” that I believed I had an obligation to incorporate the events in my teaching. Furthermore, when I suggested to my Comm. Theory class that we carry on with the projected daily schedule, they decided they WANTED to learn about theory in order to make sense of what was happening.
    • I taught on the day of the attacks, and students needed some help in giving meaning to what happened.
    • My students seemed paralyzed by the events of 9/11. We incorporated these events because no one else seemed to be talking about them. There was a tremendous need to address what was happening so students could begin to see beyond the events of the day.
  3. Instructors felt the events of 9-11 fit into the course content:
    • The events occurred shortly before my first class met, and as a relational scholar, I don’t feel we should ignore the things that affect our everyday lives in our teaching. I feel it is most powerful to use our lives in applying communication principles. As it was, we were about to discuss confirmation and disconfirmation and it seemed to me that these events could exemplify those concepts on a larger than interpersonal scale.
    • Our college is just 70 miles from New York City. Some had family or friends in the World Trade Center area. My class met on September 12. We were all stunned. A Communication Ethics class will always address such issues as news media decisions to transmit graphic images. These questions were poignantly present, and the arguments on either side very forcefully available to us.
    • I teach Epidemiology, human diseases, and environmental health. Terrorism and biological, nuclear, and chemical weapons issues are all topics that have to be dealt with in these courses. Unfortunately, now I realize that I have to make sure I teach these topics because a well-trained responder might save lives in the future. It’s sad I even have to deal with these issue.
    • After teaching a course on “Minority Images in American Media” for several years, and regularly emphasizing the manner in which the media--- especially the news media – rely on stereotypes of Arabs and others from the Middle East as terrorists, it was immediately evident to me that I needed to spend more time addressing this issue. Also, I teach a course on Visual Communication and, for several years, had difficult convincing my students that visual images communicate much more powerfully and immediately and effectively than words do. September 11 made that argument much easier to convey with the four-day bombardment of images over and over again, so including issues from Sept. 11 in my lesson plan was an obvious choice.

Together, the four themes suggested a pattern or over-arching theme among the responses – the instructors felt a need to use the course experience to work through the crisis with their students. Some did this in a more formal way by applying relevant course concepts to help critically analyze the contexts surrounding the events, others more basically sought to create a safe space for them and their students to sort through their mixed emotions.  

Grounded Theory Method. The word “theory” in this method of data analysis might be a bit confusing initially. People tend to think of theory and methods as separate, but in this groundbreaking methodology first proposed by sociologists Barney Glaser and Anselm Strauss (1967) the relationship between theory and method are more overtly celebrated. The approach has been refined and is widely practiced today in qualitative research. It is not simply a method for categorizing data, it is a method for using data categorization to reveal underlying theory (and the word theory can be used here in its most general sense, as an attempt to explain some phenomena). The resulting theory can be in the form of themes, or other attempts to explain deeper levels of meaning for what is going on in the interactions of study, referred to as substantive theory, or the result can be formal theory that will be used to advance the larger conceptual field of sociological inquiry (Glaser & Strauss, 1967). The idea is that theory can be and is most useful when it is developed from observation, rather than the other way around as done in quantitative research where the theory dictates the direction of a study. Thus Grounded Theory calls for an inductive approach to building theory from the ground up – based on research observation and analysis.

Unique characteristics. As noted above, the most unique characteristic about grounded theory is that the meanings should emerge from the data itself – from the words and behaviors of the participants. Thus, it is an excellent extension of the ethnographic approach where researchers work to honor the words and perceptions of the participants. The researcher must work to resist letting h/his perspective overly influence the meanings derived from the data. Because of this characteristic, grounded theory has become a criterion for assessing a researcher’s ethics in qualitative research and for assessing the value of the results of the study.  Such criteria demand  a highly systematic and rigorous process of data coding and meaning development.

Steps in conducting. Grounded Theory is also referred to as open coding or the constant comparative process of letting the meanings open-up -- emerge from the data, rather than imposing predetermined codes, from previous research, theory, and/or the research question(s). The terms open coding and constant comparative are actually what researchers do in steps 2-5 of the qualitative research process. They will be described further below, as they are the heart of what makes grounded theory a unique qualitative approach.

Step one (data immersion): The researcher begins with an exhaustive review of the data, trying to capture as much of it as possible.

Step two (data management):  Open coding is the initial, unrestrictive coding of data” before the researcher knows what the final categories or themes will be (Lindlof & Taylor, 2002, p. 219).  This step still requires defining the smallest unit of analysis and tracking each one closely. The researcher tries to ignore previous theory and research and let the meanings emerge through a process of constantly comparing or putting each piece of the data next to the previously coded data to look for similarities and/or differences, letting the meanings emerge piece by piece.

Open coding is a creative process. You can use a pencil, highlighter, post-it notes, or the computer to block quotes and other data and move them around as you begin to see patterns emerge. Each piece of data or instance is compared to the previous ones already categorized to see where the new piece of data belongs. Does it belong in a current category, or does it suggest a new category is warranted? You are looking for what makes sense regarding ways to organize the data. Through repeated reviews of the data the distinguishing characteristics of each category and the coding scheme will emerge.

In vivo coding.  This type of coding is done at the same time as open coding, it is a more specific type of open or grounded coding where the actual names of the codes come directly from what the participants say. The idea is to avoid the researcher imposing her/his work view as reflected in her/his labels for codes and themes. Instead the labels reflect the insider participant perspectives – it is using their words. Thus it is a preferred, more carefully grounded coding system, but it is not always possible to attain. For example, while the researcher may be quite sure the participants are engaging in a great deal of face-saving strategies, as in the case of inappropriate behaviors and creates a coding scheme for these, the participants may not realize or want to verbally recognize those behaviors.

Step three (data reduction): Eventually the number of new categories will diminish and you, as the researcher can begin to consider whether you have reached the point of what is called theoretical saturation (Glaser & Strauss, 1967)When no new categories are emerging and the categories you have created seem stable, the researcher can assume no new data is needed for now, and the analyses may be sufficient. However, if the results do not seem very insightful, rendering for example new, unique interpretations, the researcher may still decide to return to the field to collect more information.

Step four (conceptual development): When the open coding is completed (at least for now), the researcher moves to conduct what is called axial coding, basically the same general step four described previously. Think of an axis – as identifying a common framework on which the categories or themes can be located. In this step the emergent or grounded theory becomes more overt. The researcher examines how the categories created might relate to each other by conducting a metacoding – codes that attempt to connect the categories, remembering these codes must also be grounded in the data.

A useful tool to test one’s metacoding is called negative case analysis (Lincoln & Guba, 1985)The basic idea behind negative case analysis is that if the researcher’s emergent theory from the axial coding can account for a specific case, incident, person, or comment that does not seem to fit with the other data codes or categories, then the resulting theory is stronger. Sometimes researchers will return to the field to find a negative case to test their emergent theory conclusions. In this view, the negative case is not to be feared as something that will hurt the researcher’s study, but rather a piece of information that may make the resulting theory more nuanced and reflective of the complexity of participants’ lived experiences. In many cases, the information from the negative case analysis may send the researcher back to the field or back to a previous step in data analysis. The researcher must find a way to account for the negative case, which may require at an extreme completely rejecting the prior analysis, and at a minimum, more carefully describing the categories, themes, and emergent theory.        

Step five (evaluating results): As noted in the previous step, the process of assessing results has already begun. However, even after the researcher has conducted a negative case analysis and axial coding, the results are subjected to at least one more assessment. At this point the researcher may feel h/his interpretative abilities are exhausted and it is time to get a new perspective form other outside researchers, members of the community of study, or insider perspectives from some of the participants. Member checking (Lincoln & Guba, 1985) comes into play here. The researcher may choose to share the results with members of the study to see if the categories, themes, and resulting theory make sense to them. Do they represent their lived experiences? Member checking can be done very informally as in conversations with individuals, or the researcher may choose to circulate a summary of the findings with survey responses, or conduct a focus group discussion with members. It is not necessary, nor usually possible, to solicit feedback from everyone in the study.  The goal is to determine if the researcher’s conclusions ring true to the participants’ actual experiences and perceptions. It is an ethical tool to add value and credibility to the final report of the study.

(For a lecture on this topic, see sociologist Graham Gibbs, University of Huddersfield, UK, “Grounded Theory: Core Elements,” http://www.youtube.com/watch?v=4SZDTp3_New)

Discourse Analysis and Conversation Analysis

There are two other interdisciplinary methods for studying spoken and written texts or discourse, including the multiple paralinguistic ways in which the speaker delivers the words. Discourse Analysis focuses on speech (or texts) as communicative acts and examines how people use language to construct meaning. The focus of analysis is on the content of what is said and the metamessages the content conveys, perhaps about identities, relationships, positions on a social issue, etc. Conversation Analysis focuses on how the speakers communicate – the patterns of using conversational turn-taking tools such as questions, pauses, volume, and more, to negotiate identities, relationships, etc. Both methods examine the text or discourse in a line-by-line fashion with detailed transcriptions. Conversation Analysis requires the added transcription devices to indicate the amount of time between turns at talk, the paralinguistic qualities of the voice, etc. (see Gail Jefferson’s transcription system, in Sacks, Schegloff & Jefferson, 1974). As you might guess, the meanings derived from both methods are deeply embedded in the unique cultural understanding. Discourse Analysis is often used in rhetorical studies, which is reviewed in another section of this textbook. It is also used heavily in the communication sub-field of performance studies (see for example, Carlin & Park-Fuller, 2012; Palczewski, 2001). In both of these areas the researcher is attempting to make meaning of words and often nonverbal messages to provide new insights and/or greater understanding. Conversation analysis is used in some ethnography if the focus is on studying the structure of interaction in language use/conversation (see for example, Hall, 2009).

Data Write-Up

When one reaches this phase of the research process it is easy to think the creative, critical thinking work is done, but particularly in qualitative research, it is not. Writing the results is part of the meaning making process. Writing itself, is often viewed as an analysis method (Richardson, 2003). It is an opportunity for the researcher to synthesize what was learned and once again review the meanings rendered, but this time in the larger context of previous research summarized in the literature review. Often returning to this larger body of knowledge will push the researcher to make deeper and/or more complex connections among their own themes and meanings rendered.   

Not all qualitative research reports follow the traditional social science organization, but if you are not writing an autoethnography or narrative study, we recommend the social science organization for clarity. It contains four basic parts. The first is the introduction and literature review, second is a description of the research methods selected for collection and analysis, third is a description of the results or meanings formed from the data, and fourth is a discussion of the study.

A metaphor that might be helpful for visualizing how your paper will look is an old-fashioned hourglass. The introduction and literature review begins broadly and then the paper narrows to a focus on the specific study’s design. The paper ends more broadly again where the author places h/his specific results back into the context of previous research and the larger field of study to consider how the study contributed to the larger field.

There are some basic criteria for the whole report to keep in mind. They are the same criteria proposed in the data analysis process. And, by the way, these are good criteria to look for as you assess other’s reports, as well.

  • Transparency in writing about the process is key. Transparency refers to the need to be open, clear and explicit in describing the research procedures used to design the study, conduct data collection and data analysis (Seale, Gobo, Gurbrium & Silverman, 2004).The reader should be able to trace your steps and follow how you came to the conclusions you did. If the reader cannot do so, the author is hiding what may be useful information to help others understand how s/he came to the conclusions formed. 
  • Representativeness requires the researcher to be sure the results are grounded in the words and experiences of the participants and not overly controlled by the researcher’s voice. Participant quotes are your data, the narrative in the report is your voice weaving these quotes together to make meaning. Thus much of the write up will include description – the participants' actual words and behaviors to support your claims (themes or conclusions). The reader should not feel the patterns, themes, or other conclusions were drawn out of thin air. The reader should be able to see 2-3 exemplars for each claim made about the results of the study.
  • Reflexivity is related to transparency. It is an effort to be open with the reader and acknowledge that you as the researcher are a part of what is being researched. You are not an objective observer uninfluenced by the study or detached from its results. The reflexive author is careful to separate h/his own feelings and comments from what the participants said, and careful to note when such a clear separation is not possible. Reflexive writing is being critical of one’s own work. This does not mean always being negative, pointing out limitations, just that one recognizes h/his potential role in the process.

Introduction and Literature Review

The introduction and literature review are combined in this step. It is useful to think of them as a coherent rationale for the study you conducted or propose to conduct. The Rationale is drawn from life experiences, review of news reports, popular press, secondary sources (as in websites, textbooks or published reviews of literature and most importantly, previous original scholarship on the topic. As in quantitative research, research questions that guide original qualitative research are generally expected to be based on previous research. Some ethnographers and autoethnographers prefer not to review the literature prior to collecting original research so that their views are not tainted. In this case, the literature view portion of the final report would go in the discussion of the results of the study, rather than as a part of the rationale of the study.

The rationale is built by answering the following questions in the paper, usually in this order:    

  1. A statement of the problem that needs to be studied (this can be a social problem, an academic theory that needs to be developed, a lack of experience needed, etc.) The definition of problem is meant to be broadly defined here. This is the attention-getter of the paper. It should compel the reader to want to know more. (Usual length is from a paragraph to two pages, double-spaced.)
  2. A statement of why this problem needs to be addressed – offering evidence to show it is important and that it can be addressed through original research. (Usual length is from a few paragraphs to a few pages, double-spaced).
    • Embedded in the statement of the problem and why it needs to be addressed, the researcher introduces for the first time any key terms that will be used repeatedly in the study. These should be briefly defined for the reader with a citation of the source. For example, if you were doing a study on how people build an intercultural relationship with someone from a different race or ethnicity, What key terms would you be using? Mostly likely you would need to define culture, race/ethnicity, prejudice, intercultural relationships and relational development. While there may be other concepts, only the most central ones need to be introduced here.
  3. A preview of the type of research warranted from the review of the problem.
  4. A review of previous research.
    • There are multiple ways to organize a review of the literature. You will want to select what helped you organize the information and understand it best. Topical organization is most popular, but it might be useful to organize the review in a chronological fashion. The organization of sub-headings should help lead the reader to understand why you have decided on the design of the study you selected.
    • The question that is commonly asked is, “how much do I need to include from each study?” There is no specific answer – thank goodness. There is room for your thoughts. Basically include what is most relevant to the present study you will or did do and not more. Thus, if you adopted methods from a study then describe the study’s methods, but if you are only citing the study as evidence that your topic is important, then just cite the study, as in parentheses at the end of a statement you make as in: (Ellis, 2004). 
    • A review of literature is not just a cut and pasting of a series of research abstracts. Your voice synthesizing the previous research should be central in the review.
  5. The review of previous research literature usually concludes with the proposal of one to three overall research questions the author will attempt to answer with the specific study proposed. (Length varies a great deal depending on the instructor’s assignment and/or publishers’ page limits. A rule of thumb is that it be almost as long as the reporting of results from your study, from five to sixteen pages).

Research Methods

As in quantitative research, there are two sub-headings to this: data collection and data analysis.

Data Collection. Here the researcher describes the general research design proposed to answer the research questions stated at the end of the literature review. The criterion of transparency is key to follow here. This section may range from two to approximately six pages. This section includes the following, usually in this order:

  • Participants – method used for soliciting participants, ethical guidelines followed in soliciting them and in terms of promises of confidentiality, or other concerns to not ask more of them than is necessary. Forms of consent are referenced and included in the appendices of the paper. Lastly, in qualitative research, if there are not too many participants, the researcher might include their first name or pseudonym and a brief description of each person’s demographics collected as relevant to the study.
  • Data Collection – a detailed description of the methods chosen (e.g. interviews, ethnographic observation, …), how they were conducted, and why these choices are relevant for the general research questions asked. Actual interview questions are usually put in the appendices.

Data Analysis. Here the researcher explains which analysis tools (e.g. content analysis, theme analysis) were selected and why they are a good fit for the general research questions asked and the specific data collection methods employed. This section includes:

  • Description of how data immersion was accomplished (e.g. length of time taken, number of readings, etc.)
  • Definition of the unit of analysis used with a few examples
  • How the data was coded and what emerged as the coding scheme
  • How the researcher managed the data and reduced it to a focus that was attainable and relevant

Results

This is where the author describes what was constructed in step four and five of the general data analysis process: the conceptual development and evaluation of results or meta-analysis. The author describes what meanings were taken from the data based on the analysis process. A suggested outline:

  • A list of the categories or themes and exemplar texts for each to demonstrate them
  • If negative case analysis was conducted, how it changed the resulting interpretations.
  • Results from the meta-analysis of themes or categories. What did the researcher find about how these might be related? What is the deeper and broader level of meaning attained from the analysis process?

This is where a criterion used in autoethnography might be useful (Ellis, 2004). The author is expected to use evocative writing -- writing that calls up emotional responses from readers. It compels readers to engage with the material beyond a cognitive level. By calling for writing that is compelling, we are not suggesting the author try to manipulate the readers’ emotions. It is just that in autoethnography the quality of a study is measured so to speak by how much it compels the readers, draws them into the reality being described, and helps them see the experience from an inside view. We think this is a good criterion for most qualitative research on human experience.

Conclusion

If you imagine the introduction and conclusion of your research report as covers on a book that are mirror opposites, the conclusion should respond to and connect with the introduction. The goal is that when the reader finishes s/he will feel like the book just snapped shut. The author connects the findings from the present study to the larger social problem being addressed and the larger body of research in the literature review. When the covers snap shut, the author has done a good job of answering the basic question of, “so what?”  “Why these research questions, why these participants, why this research design, why these interpretations of meaning, why do the results matter?” The following outline is offered to help assure you answer the question of “so what?”

  • Preview of the sub-headings in the conclusion
  • Summary of the key findings (themes, theory development and sub-categories supporting them  (less quotes are needed here).
  • Contributions of the study – both practically (e.g. for practitioners working with students, patients, for people to apply in their daily life, etc.) and academically (e.g. contributions to theory, methods, topic of study – the larger field of literature on the subject). Note this is especially where the author returns to the literature review and selects specific sources that the results of the study might speak to.
  • Limitations of the study (e.g. in participant selection, data collection and analysis methods applied.
  • Suggestions for future research (often based on the limitations of the present study).

Other Aspects of the Paper

For a more detailed description of the writing process and format, also see Chapter 7 of this book: “Conclusion: Presenting Your Results".

There are other parts of the academic paper you should include in your final write-up. We have provided useful resources for you to consider when including these aspects as part of your paper.

For an example paper that uses the required APA format for a research paper write-up, see the following source: http://owl.english.purdue.edu/media/pdf/20090212013008_560.pdf.  However, this example uses quantitative data, so be aware the results section will not look like your qualitative study results. Qualitative reports tend to be longer.

Abstract & Titles.

http://writing.wisc.edu/Handbook/presentations_abstracts.html 
http://owl.english.purdue.edu/owl/resource/560/1/

Tables, References, & Other Materials.

http://owl.english.purdue.edu/owl/resource/560/19/ 
http://owl.english.purdue.edu/owl/resource/560/05/

Data Presentation

Instructors will often ask you to present an oral version of your study in addition to the written report. This is a great way to help you gain more full comprehension of what you did in your study. If you can explain it well to others, you likely understand more deeply what you accomplished. The oral presentation is also a great way to prepare you for academic or other professional conference presentations that will help add to your resume. Prior to submitting original research to be considered for publication in journals or edited books, researchers share their studies orally to gain feedback and revise the written version of a study for submission to publication.

Two of the most common venues are oral presentations as a part of a panel of speakers and poster presentations. You might also be called upon to write an executive summary of the results of your study that makes it more accessible for people to review quickly. There are good resources for doing all of these online, so we have provided these here.

Oral Presentations

http://writing.wisc.edu/Handbook/presentations_oral.html 
http://writing.wisc.edu/Handbook/presentations_delivery.html

Poster Presentations

http://writing.wisc.edu/Handbook/presentations_poster.html

Executive Summary

http://www.csun.edu/~vcecn006/summary.html
http://www.stanford.edu/group/gender/ResearchPrograms/DualCareer/DualCareerFinalExecSum.pdf (an example of an executive summary for university policy makers, from a research study on dual career academic couples)
http://www.kff.org/entmedia/upload/7618ES.pdf (another example of an executive summary from a study of food advertising to children on television)

References

Alhojailan, M.I. (2012). Thematic analysis: A critical review of its process and evaluation.  West East Journal of Social Sciences, 1(1), 39-47. [access at http://www.westeastinstitute.com/journals/wp-content/uploads/2013/02/4-Mohammed-Ibrahim-Alhojailan-Full-Paper-Thematic-Analysis-A-Critical-Review-Of-Its-Process-And-Evaluation.pdf]

Blumer, H. (1986).  Symbolic interactionism: Perspective and method. Berkeley: University of California Press.

Bochner, A., & Ellis, C. (1992). Personal narrative as a social approach to interpersonal communication. Communication Theory, 2, 65-72.

Carlin, P. S., & Park-Fuller, L. M. (2012). Disaster narrative emergent/cies:
Performing loss, identity and resistance. Text and Performance Quarterly, 32(1), 20-37.

Cohen, M., & Avanzino, S. (2010). We are people first: Framing organizational assimilation experiences of the physically disabled using co-cultural theory. Communication Studies61(3), 272-303.

DeFrancisco, V., and Chatham-Carpenter, A.  (2000). Self in community: African American women’s views of self-esteem.  Howard Journal of Communication, 11 (2), 73-92.

Denscombe, M. (2010). The good research guide for small-scale social research projects (4th ed.). Maidenhead, England: Open University Press.

Denzin, N. K. (2006). Sociological Methods: A Sourcebook. Piscataway, NJ: Transaction.

Dilthey, W. (2010). Selected works, volume II: Understanding the human world. R. A. Makkreel & Frithjof, R. (Eds.). Princeton, NJ: Princeton University Press.

Ellis, C. (2004). The ethnographic I: A methodological novel about autoethnography. Walnut Creek, CA: AltaMira Press

Ellis, C., & Bochner, A. (2000). Autoethnography, personal narrative, reflexivity: Research as subject. In N. Denzin & Y. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 733-768). Thousand Oaks, CA: Sage.

Gadamer, H. G. (1976). Philosophical hermeneutics (D. E. Linge, Trans.). Berkeley: University of California Press.

Geertz, C. (1973). The interpretation of cultures. New York: Basic Books.

Glaser, B., & Strauss, A. (1967). The discovery of grounded theory: Strategies for qualitative  research. New York: Aldine.

Grbich, C. (2007). Qualitative data analysis: An introduction. Thousand Oaks, CA: Sage.

Hiles, D. R. (2008). Transparency. In L. M. Givens (Ed.), The Sage encyclopedia of qualitative  research methods. Thousand Oaks, CA:Sage.

Hundley, H. L., & Shyles, L. (2010). US teenagers’ perceptions and awareness of digital technology: A focus group approach. New Media & Society12(3), 417-433.

Husserl, E. (1990). On the phenomenology of the consciousness of internal time (1893–1917), (J. B. Brough, Trans.). Dordrecht: Kluwer.

Keyton, J. (2011). Communication research: Asking questions, finding answers (3rd ed.).
New York: McGraw Hill.

Lincoln, Y, S., & Guba, E. G. (1985). Naturalistic inquiry. Beverly Hills, CA: Sage.

Lindlof, T. R. (1995). Qualitative communication research methods. Thousand Oaks, CA: Sage.

Lindlof, T. R., & Taylor, B. C. (2002). Qualitative research methods (2nd ed.).
Thousand Oaks, CA: Sage.

Lofland, J. & Lofland, L. H. (1995). Analyzing social settings: A guide to qualitative observation and analysis. Belmont, CA: Wadsworth.

Meares, M. M., Oetzel, J.G., Torres, A., Derkacs, D., & Ginossar, T. (2004). Employee mistreatment and muted voices in the culturally diverse workplace. Journal of Applied Communication, 32 (1), 4-27.

Owen, W. F. (1984). Interpretive themes in relational communication.  Quarterly Journal of Speech, 70, 274-287.

Palczewski, C. H. (2001, Summer). Contesting pornography: Terministic catharsis and definitional argument. Argument and Advocacy, 38, 1-17.

Richardson, L. (2003). Writing: A method of inquiry.In N. K. Denzin & Y. S. Lincoln (Eds.), Collecting and interpreting qualitative materials (pp. 499-541). Thousand Oaks, CA: Sage.

Ricoeur P. (1981). Hermeneutics and human science: Essays on language, action and interpretation (J. Thompson, Trans.). London: Cambridge University Press.

Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the organization of turn-taking in conversation, Language, 50, 696-735.

Saldana, J. (2009). The coding manual for qualitative researchers. Thousand Oaks, CA: Sage.

Schutz, A. (1967). Phenomenology of the social world. (G. Walsh & F. Lehnert, Trans.). Evanston: Northwestern University Press.

Schwandt, T. (2001). The SAGE dictionary of qualitative inquiry. Thousand Oaks, CA: Sage.

Seale, C., Gobo, G., Gubrium, J. F., & Silverman, D. (Eds.). (2004). Qualitative research practice. London: Sage.

Spiggle, S. (1994). Analysis and interpretation of qualitative data in consumer research. Journal of Consumer Research, 21, 491-503.

Stewart, D. & Mickunas, A. (1990). Exploring phenomenology: A guide to the field and its literature. Athens, OH: Ohio University Press.

Weber, M. (1962). Basic concepts in sociology. (H.P. Secher, Trans.) New York, Citadel Press.

Weinstein, S. (2007). Pregnancy, pimps, and “clichèd love things”: Writing through gender and sexuality. Written Communication, 24(1), 28-48.

Wieder, D. L (1999). Ethnomethodology, conversation analysis, microanalysis, and the ethnography of speaking (EM-CA-MA-ES): Resonances and basic issues. Research on Language and Social Interaction, 32(1-2), 163-171.

Wong, S. & Goodwin, R. (2009). Experiencing marital satisfaction across three cultures: A qualitative study. Journal of Social and Personal Relationships, 30, 734-749.