Abstract
Historically, bias has been viewed as a problem with qualitative research, in large part because the researcher is often a qualitative study’s primary (or, in some cases, only) research “instrument.” Even most constructivists who reject traditional notions of scientific objectivity want to produce reasonably accurate reconstructions of their research participants’ interpretations of the social world. In fact, presenting accurate reconstructions is one thing constructivists mean when they talk about a study’s trustworthiness. Various procedures have been developed to minimize bias and ensure trustworthiness. The Formative Influences Timeline (FIT) is one such procedure. This article provides a brief history of bias in qualitative research, describes the FIT and how to use it, and reviews nonprofit research studies that either employed—or could have employed—the FIT to produce data relatively uncontaminated by researchers’ a priori assumptions. The article concludes by acknowledging limitations of the FIT.
Journals in the field of nonprofit and philanthropic studies are increasingly publishing qualitative research (for prominent examples, see AbouAssi, 2012; Academy of Management Journal, 2009; Bies, 2010; Carnochan et al., 2014; Dale, 2018; Englert et al., 2020; Kissane & Gingerich, 2004; Knutsen & Brower, 2010; McNamee & Peterson, 2016). Qualitative research, however, presents challenges for journal editors and reviewers, as well as for research consumers. One rather daunting challenge relates to the problem of bias.
This problem of bias stems, in part, from the fact that, in qualitative studies, researchers, themselves, are often the primary—and, in some cases, the only—research instrument. In this article, we discuss the Formative Influences Timeline (FIT), a very simple instrument that minimizes the likelihood of researcher influence on qualitative interview results. First, we take a brief historical look at the issue of bias in qualitative research. Second, we describe how the FIT addresses the problem of bias in qualitative interview studies and provide step-by-step instructions for implementing the FIT procedure. Third, we discuss two studies in which the FIT strategy unearthed data that were largely uncontaminated by researchers’ a priori assumptions; we also discuss a study in which the FIT could (and probably should) have been used. Finally, we acknowledge limitations of the FIT.
Qualitative Research and the Issue of Bias
Qualitative Methods in Anthropological Research
Ethnography
The origins of qualitative methods as social science research strategies can be traced back to the first quarter of the 20th century and, more specifically, to cultural anthropologist Bronislaw Malinowski’s research in Melanesia. When Malinowski (1922/2013) arrived in Melanesia, he discovered a world that was radically different from the society that socialized and formally educated him. The people whose culture he intended to study spoke a different language, engaged in what, for him, seemed like alien activity, and had views of the world that diverged, in fundamental ways, from his own.
It quickly became apparent that Malinowski’s training at the London School of Economics could not be directly applied to the culture he was intent on learning about, and, over time, largely through trial and error, Malinowski developed the methodology he dubbed participant observation. Years later, after Malinowski returned to the London School of Economics with a chair in anthropology, one of his students, Hortense Powdermaker (1967), wrote a book with a title that succinctly captured the duality of the participant observer role: Stranger and Friend: The Way of an Anthropologist.
Participant observation became the methodological centerpiece of ethnography, a term that refers to what most cultural anthropologists do, even today: Cultural anthropologists systematically study and describe cultures and the worldviews of the individuals who are members of an organization of one sort or another.
Bias in ethnographic research
A concern about bias has always loomed large over ethnographic work because, even when an anthropologist played what Powdermaker (1967) called the stranger role in the stranger/friend dichotomy to analyze and critically reflect on the evidence gathered and the interpretations this evidence support, the anthropologist normally does this work alone. Consequently, one cannot check for intersubjective agreement, just as there normally is no reliable and validated instrumentation to guard against the intrusion of the anthropologist’s personal bias. Ethnographers normally are the primary (and, often, the only) “instrument” employed in ethnographic studies.
Concern about bias in ethnographic work became especially acute in 1983 when Freeman published Margaret Mead and American Samoa: The Making and Unmaking of an Anthropological Myth. Freeman had attempted to replicate 1 the famed anthropologist’s influential work on the sexual attitudes and practices of American Samoan youth reported in Coming of Age in Samoa: A Psychological Study of Primitive [sic] Youth for Western Civilization (Mead, 1930/2001). He concluded that Mead had entered the field with an a priori storyline in mind and subconsciously made the data she gathered fit what she expected to see and hear. Freeman also argued—and provided some evidence to support the claim—that Mead’s research participants intuited all of this and, basically, told her what they knew she wanted to hear.
Today, there are still different verdicts about whether Mead’s account of sexual practices in America Samoa is mostly accurate or merely a figment of Mead’s fertile imagination. Even scholars’ review of Mead’s field notes preserved in the Library of Congress could not end the controversy. For a detailed account of the Freeman/Mead controversy, see Shankman’s (2009) The Trashing of Margaret Mead: Anatomy of an Anthropological Controversy.
Bias and Qualitative Research in Other Fields of Study
Sociological research
Anthropology was not the only field to employ qualitative methods during the 20th century, and bias often was a concern in these other fields, as well. In the discipline of Sociology, for example, Glaser and Strauss (2017, c1967) attempted to build theory qualitatively and inductively; “grounded theory,” they called it. Also, sociologists who called themselves either symbolic interactionists (Blumer, 1969) or ethnomethodologists (Garfinkel, 1967) assumed that social action is rooted in the meaning human beings attach to social phenomena and that human beings construct—and constantly reconstruct—meaning through social interaction. Their goal was to explicate, once again, inductively, the social interaction processes through which meaning is made and constantly reconfigured.
In all these inductive versions of qualitative sociological research, it was not feasible to employ procedures such as hypothesis testing or experiments with control and treatment groups. These, of course, were (and are) the sorts of procedures traditional sociologists and other social scientists rely on to support claims about the validity of their findings.
Qualitative methods in applied fields
Toward the end of the 20th century, journals in applied fields—fields like business administration and nonprofit/philanthropic studies—began publishing qualitative studies. In 2009, for example, the editors of the Academy of Management Journal (AMJ) wrote the following in an editorial: Over the past several years, we at AMJ have worked diligently to increase the number and quality of the qualitative research papers we review and publish . . . Our efforts to increase high quality qualitative work in the journal stems [sic], in part, from our mission to publish research that has the highest impact. Qualitative research certainly fits this bill. (p. 856)
Once qualitative research made its way into applied fields of study and onto the pages of these fields’ premier academic journals, interviewing had become the primary qualitative data collection method. This occurred primarily because it was not feasible for most researchers in applied fields to engage in extended fieldwork. There were other differences, as well, some of which involved rethinking traditional concerns about bias.
Rethinking the bias issue
In fact, some qualitative researchers at the time went on the offensive and turned traditional social scientists’ critiques about bias in qualitative research on their head. In 1986, for example, feminist qualitative researcher Patti Lather published a paper titled “Issues of Validity in Openly Ideological Research: Between a Rock and a Soft Place (Lather, 1986a).” This title’s not-so-subtle subtext (and the argument articulated quite explicitly in the actual text of the article) was that all empirical research is ideological. Before researchers can count, for instance, they must determine what to count and how to characterize what they are counting. The language they select to formulate their research questions and report their “empirical” findings—whether, for example, they opt to refer to their research subjects as culturally disadvantaged or, simply, as culturally different—has a profound impact on the policy and practice implications of their studies.
In short, from Lather’s perspective, empirical research of any kind is never completely empirical. However, the problem of bias is less insidious in qualitative research where the influence of researcher subjectivity is obvious. By contrast, in traditional social science, the subjective or ideological elements of a study normally are neither acknowledged by researchers nor recognized by research consumers.
Qualitative Inquiry and Research Paradigms
The positivist and constructivist paradigms
Arguably, the most influential book during the period when qualitative methods began to be used in applied fields was Lincoln and Guba’s (1985) Naturalistic Inquiry. The authors argued the use of qualitative methods was not simply about a choice of methods; rather, the emerging interest in qualitative methods was symptomatic of something analogous to the sort of paradigm revolutions—for example, the shift from Newtonian to Einsteinian physics—that Kuhn (2012, c1962) had documented in the physical sciences. Qualitative researchers, Lincoln and Guba claimed, normally embrace a constructivist ontology and an interpretivist epistemology and reject traditional ways of working, ways the authors referred to as “positivism.”
Lincoln and Guba’s alternative to positivism was a new research paradigm they initially called naturalistic and later redubbed the constructivist paradigm. The naturalistic/constructivist paradigm included, among other things, new vocabulary with which to talk about the issue of bias. In the new paradigm, for example, talk of validity was replaced with talk about a study’s trustworthiness, and the term “triangulation” replaced traditional researchers’ concerns about reliability.
These terminology changes signaled more than changes in wording. In her now classic paper “Why Triangulate?” for example, Mathison (1988) noted that the term reliability refers to measures of consistency, but, when researchers assume knowledge is constructed, they should not expect findings to necessarily be consistent. Findings generated through different research methods or from interviewing different people who may have different life experiences or play different roles in an organization might be consistent with each other, but they also may diverge and may even be contradictory. A lack of consistency is not a problem for constructivists who engage in triangulation, Mathison argued; it simply represents a challenge for the qualitative researcher to make sense of differences.
Of course, not all qualitative researchers believed they were part of the paradigm revolution Lincoln and Guba claimed was occurring. Positivist qualitative researchers consciously attempted to eliminate, or at least minimize, the sort of bias that concerned traditional social scientists, even though they could not normally use the same procedures for doing this that quantitative researchers employed.
Research paradigms in nonprofit/philanthropic studies
Most of the studies published in the nonprofit field’s major journals, for example, appear to be more reflective of the positivist qualitative research tradition than Lincoln and Guba’s constructivist paradigm. This orientation is, almost certainly, not the result of journal editors and reviewers consciously rejecting Lincoln and Guba’s conception of constructivism. Indeed, most journal editors and reviewers are likely agnostic when it comes to ontological and epistemological questions. Like most members of the research community, they likely spend their time doing and/or reviewing empirical studies rather than engaging in abstract debates about the ontological and epistemological foundations of the studies with which they are engaged. But, because the nonprofit field has been influenced by academic disciplines such as Public Administration, Business Administration, Economics, and Sociology, fields of study that, historically, have emphasized the generation and, especially, the testing of theory, it should not be surprising that even scholars who use qualitative methods have tended to implement approaches to qualitative research that lean toward the positivist end of a positivist/constructivist continuum.
This leaning is quite evident in work that employs Lincoln and Guba’s language to describe the methods that were used in a study but that use those methods to accomplish purposes that Lincoln and Guba almost certainly would find questionable. Bies’ 2010 paper, “Evolution of Nonprofit Self-Regulation in Europe,” for instance, contains the following statement in the paper’s methodology section: To increase the rigor of the study, researchers engaged in a process of verification, by conforming with several strategies specific to case study research as defined by Creswell (1998), including member checking, prolonged interaction, and triangulation of data.” (p. 1060)
Similarly, Carnochan et al. (2014) wrote the following in the methods section for their study of four nonprofit human services organizations: “This qualitative study involved interviews, focus groups, and review of archival records. The use of multiple sources of data allowed for data triangulation and enhanced the internal validity of the findings” (p. 1018).
Given Carnochan et al.’s emphasis on using triangulation to demonstrate internal validity, a concept associated with the positivist research tradition, and Bies’s (2010) use of triangulation for “verification” purposes, the authors did not view triangulation in quite the same constructivist way Mathison (1988) did. Rather, in these studies, triangulation seems to be a qualitative version of quantitative researchers’ procedures to assess reliability. 2
There is nothing inherently wrong with this, of course. Most people now understand that paradigms are not right or wrong; rather, they help us play different research “games” and accomplish different research purposes. The authors’ conception of triangulation, however, does signal that the use of qualitative methods has tended to play out a bit differently in the nonprofit field than in certain other academic fields of study.
In some other fields of study, in fact, Lincoln and Guba’s paradigm talk has opened the door for a plethora of new research paradigms and, in some instances, quite different research methods, for example, Harding’s (2005) and Lather’s (1986b) discussions of feminist epistemology and methods; Wilson (2008) and Tuhiwai Smith’s (1999/2012) discussions of the Indigenous paradigm and decolonizing methodologies; and Dillard’s (2000) and Wright’s (2003) discussions of an epistemology rooted in African American women’s life experiences. Unfortunately, a detailed discussion of paradigm proliferation and its implications for nonprofit and philanthropic studies is beyond the scope of this article.
The Bottom-Line About Bias
This article, after all, has a more circumscribed purpose: to discuss the FIT, a strategy we have found useful for minimizing researcher bias in qualitative interview studies. Here it is enough to note that the FIT can be used in both positivist qualitative research and in constructivist forms of qualitative inquiry, as well.
After all, most constructivists still accept objectivity as a regulative ideal. Philosopher of science, Sandra Harding (2005), for example, coined the term “strong objectivity” to refer to research that both (a) acknowledges the inevitable impact of a researcher’s personal history, methods, language, positionality, and socialization and (b) uses researcher reflexivity to expose, manage, and, to the extent possible, minimize the impact of subjective influences.
Furthermore, even the most committed constructivists still want to produce reasonably accurate reconstructions of their research participants’ interpretations of the social world. This is basically what constructivist qualitative researchers are referring to when they talk about a study’s trustworthiness. Indeed, the methodological procedures Bies (2010) associated with the “process of verification” in her study were first articulated in Lincoln and Guba’s (1985) discussion of how to ensure the trustworthiness of studies conducted within their constructivist paradigm.
The FIT: An Interview Strategy to Minimize Bias
The goal of the FIT interview strategy is to systematically identify what research participants believe to be the developmental influences on whatever is being studied. It does this without researchers dictating, through their questioning, the specific topics research participants focus on and discuss during an interview. We (and our graduate students) have used this strategy to explore motivations for philanthropic giving, evaluate nonprofit programs, and identify the evolution of antiracist activism (Pearson-Dawe, 2017).
The strategy is especially useful for documenting perceptions of change over time, either in individuals or in organizations or programs. In fact, one of us developed the strategy nearly 20-years ago while serving as a consultant for a foundation involved in a public/private partnership with a state governor’s office. The governor’s office was overseeing the implementation of the governor’s new literacy initiative in the state’s elementary schools, and the foundation agreed to conduct a formative evaluation of the initiative and its impact. The author was hired by the foundation to head a team of researchers and manage the evaluation process.
The evaluation primarily employed a multilevel modeling design to document the initiative’s impact on student achievement, but the governor’s office also wanted to know about any problems school personnel encountered while implementing the initiative. The team of researchers employed the FIT to respond to this concern, and it proved invaluable for understanding the often quite different ways different schools approached implementation and the different problems they encountered during the first year of the initiative.
FIT Procedures
A FIT is typically conducted at the beginning of an interview where it serves as both a data collection protocol and an opportunity to develop rapport by listening as interviewees tell their stories, largely uninterrupted by the researcher. The length of time required to conduct a FIT varies. We have conducted a FIT in as little as 15 and as much as 90 minutes. We suggest that researchers communicate anticipated timing to participants in advance so participants can adjust the pace accordingly.
Prior to conducting a FIT, researchers must prepare, for each interviewee, a blank sheet of paper laid out horizontally with a single horizontal line down the center of the page. This line will be described to the participant as a timeline of their life or portion of their life, if the focus is on personal development, or the life of an organization or program, if the study has an organizational/programmatic focus.
The FIT interview, itself, can be conducted either in person, on the phone, or via SKYPE-like social media. If the FIT interview is conducted in person, researchers should ensure they have a quiet and confidential space in which to conduct the interview. In the case of phone or electronic interviews, the timeline document should be mailed or emailed in advance.
At the start of the interview, participants are given the sort of instructions below that have been adapted from a recent interview protocol: Imagine the line on this paper is a timeline of your life [or of the organization/program over time]. I would like you to identify the events and memories that you believe shaped who you are today as a [philanthropist, volunteer, board member, etc.] [or what your organization/program is like today]. Please take about five minutes and jot down notes about these events. These notes are not for me. You will use these notes to prompt your memory as you describe these events and memories to me.
After the interviewee completes the timeline and finishes jotting down notes, the research says something like the following: Now, please describe each of the events and memories you’ve identified as influencing who you are today as a [philanthropist, volunteer, board member, etc.] [or what your organization or program is like today]. I may stop you at certain points to ask for clarification or more details.
As an interviewee describes events, the researcher may ask for additional information about entries on the timeline. The researcher might also ask whether there were other influences than the ones the interviewee put on the timeline. Depending on the type of qualitive research one is conducting (e.g., positivist or constructivist), the researcher may be guided in their probing by (a) the study’s research questions, (b) the literature review for the study, or (c) topics that emerge naturally as part of the conversation. The goal, however, is always for the interviewee to take the lead and for the interviewer’s probing to be a response to what the interviewee has been saying.
The FIT is a data collection strategy. How data collected via the FIT are used in data analysis will depend on the purpose of the study in which the FIT strategy is employed. With permission from the interviewee, the researcher can retain the actual timelines constructed by study participants for use during data analysis, though most of the data to be analyzed will be contained in the audio recordings and/or interviewer notes. Of course, examples of participants’ timelines could be displayed in the research report, though we have not done this.
Minimizing Bias
As was already suggested, the main advantage of using the FIT strategy is its ability to minimize a researcher’s influence during interviews while still providing a clearly defined “choreography” qualitative researchers can employ to ensure interviewees’ time is used efficiently. Virtually any predefined interview protocol is likely to ensure the efficient use of interview time, of course, but most interview protocols tend to limit the answers interviewees provide, and they also have the potential to signal interviewees the story the researcher expects and/or wants eventually to tell. This signaling, in turn, may “lead the witness” and skew the information interviewees provide. After all, even if Freeman (1983) got it wrong and the Samoan youth Margaret Mead interviewed did not tell her what they intuited she wanted to hear, the social desirability phenomenon has been well documented in other social science studies (Fisher, 1993).
By contrast, when the FIT is employed, the substance of interviews is dictated almost entirely by research participants, not by the researcher. The items participants place on their timelines determine both what is talked about during the interview and any follow-up questions the researcher might ask. In short, each study participant has a great deal of autonomy in terms of what and how to present their response to the researcher’s request to draw and talk about formative influences either on themselves or in an organization or program they know about.
Comparisons
The FIT is not the only interview strategy that allows for the collection of data that are relatively uncontaminated by researchers’ questioning. Arguably, one of the best-known set of procedures that also do this is detailed in Spradley’s (1979/2016) The Ethnographic Interview. In fact, Spradley’s grand tour questions (i.e., questions that ask the interviewee to provide either a literal or metaphorical journey around a place or series of events) are similar to the timeline task at the center of FIT interviews.
Spradley’s procedures, however, are far more complicated than the FIT procedures outlined above. Indeed, asking grand tour questions is simply the starting point for a relatively complex, multistage, and symbiotic process that alternates data collection with data analysis. Spradley, after all, was a cognitive anthropologist, and, as such, his goal was to map the cognitive structures in his research participants’ heads by documenting the substance and structure of their language.
One of us, years ago, employed Spradley’s methods in a study funded by the Rockefeller Brothers’ Fund that focused on an organization in which members had developed a specialized vocabulary (Donmoyer, 1985). Adapting Spradley’s methods to ferret out the meaning of organizational members’ native language proved quite useful in that study, but the author who conducted that study has not found Spradley’s data collection and analysis techniques to be a good fit (no pun intended) in subsequent nonprofit studies. Other nonprofit researchers who are not cognitive anthropologists may also find Spradley’s methods inappropriate for most of the studies they conduct.
Of course, one could argue that what can be achieved using the FIT could just as easily be achieved through a protocol-free conversational interview (Patton, 2015) conducted by a skilled interviewer who knows how to ask open-ended questions. This claim is probably correct, but the key words in the previous sentence are “skilled interviewer.” Traditionally, instruments, even relatively simple ones like the FIT, have been used in social science research because we cannot count on all researchers to be skilled interviewers. In addition, proposal writing is considerably easier if there is an instrument to write about and include in the appendices of the proposal, even if the instrument is as simple as the FIT. One of us works at an institution where the Institutional Review Board (IRB) is skeptical of conversational interviewing, but board members have had no difficulty approving proposals for studies that employ the FIT.
Examples
In this section, we present two examples of the FIT being used to generate data that led to more comprehensive and trustworthy findings. Then, we present an example of one study in which a FIT was not used but, in hindsight, could have contributed positively to the study.
Example #1: When Measuring Outcomes is Not Just About Measuring Outcomes
This first example is from a mixed-methods study examining how philanthropists think about their giving (Jones, 2018). The research questions were as follows: (a) what are the action logics (i.e., developmental levels) of the donors included in the study, (b) how, if at all, does a donor’s action logic appear to influence philanthropic decisions, and (c) to what extent does a donor seek feedback and/or engage in self-reflection about his or her philanthropic practices, and does the answer to this question appear to be related to a donor’s action logic? Each participant engaged in two interviews, the first of which was a 60- to 90-minute FIT. Here, we describe data collected during Joseph’s 3 FIT.
During the interviews, it became obvious Joseph had strong ideas about the importance of measuring nonprofit outcomes. In fact, he believed organizations that do not have demonstrable, replicated outcomes should be eliminated. On the surface, this approach seems reasonable, particularly if one knows Joseph is a businessman. However, the FIT revealed a possible deeper motive for Joseph’s adamant commitment to demonstrable, replicable outcomes that helped make sense of the emotionally charged conflicts Joseph had experienced with multiple organizations that failed to meet his expectations.
During a 75-minute FIT, Joseph characterized the early stages of his career as tumultuous. He recalled a time in his early 20s when his father said something to the effect of, “You’ll never amount to anything. I’ll have to hire you because no one else will.” Joseph somewhat cavalierly stated he had been proving his father wrong ever since. He went on to describe several situations where he had to walk away from supporting nonprofit organizations that were not managing the organization according to his standards.
Because of the insights generated during the FIT interview, we began to see Joseph’s laser-like commitment to producing measurable outcomes as potentially emanating from something more than just a concern with an organizational success. His near obsession with measuring outcomes—an obsession that continually undermined his relationship with the organizations and organizational leaders he was trying to help—could reasonably be interpreted as a legacy of his attempts to prove his father wrong throughout his life. In short, understanding how his father influenced his development as a philanthropist was a clue that helped us understand how and why he acted as he did, even when his actions seemed counterproductive. It is unlikely that we would have thought to ask about Joseph’s family history during the more traditional interview conducted after the FIT had we not first employed an interview strategy that let interviewees tell their own developmental stories.
Example #2: When What Looks Like Antiracist Action is not Actually Antiracist
The second example comes from a primarily qualitative study conducted by one of our graduate students examining the experiences of white people who volunteered their time to actively dismantle systemic racism (Pearson-Dawe, 2017). The research questions were as follows: (a) to what extent, if any, are the participants aware of their own white privilege, and (b) what factors, if any, are related to a white person’s willingness to dismantle white privilege? Participants were identified through a referral sampling method, the goal of which was to select study participants engaged in antiracist activities.
Claire, a Caucasian senior, was the oldest participant in the study. Through a 20-minute FIT, the researcher realized that what seemed to be Claire’s antiracist actions was not actually undergirded by antiracist motives. Specially, Claire was asked to describe the events and memories that led her to become an antiracist activist; however, the events and memoires she recalled demonstrated a lack of awareness of antiracist intent and, in fact, exposed what could be characterized as racist attitudes. For example, she described how she, a white person, volunteered to lead an event of a state-level chapter of the NAACP. She bragged about taking over the event and making it better than it had been when a person of color had run it.
This memory raised a red flag for the researcher, and after reviewing the interview transcript, what became apparent was what Claire did not share during the FIT. The other participants in this study all described an evolution of understanding of their white privilege. They described experiencing discomfort related to their privilege, instances where they had resisted antiracist action or even engaged in backtracking, and how they paid attention to the way they wielded their whiteness (intentionally or not) in various spaces.
This sort of nuanced discussion of white privilege was nowhere to be found in Claire’s transcript. Claire’s understanding of white privilege was not nuanced; nor had it changed over time. Instead, Claire stated that she had learned about slavery in grade school, and this single elementary-school lesson appeared to shape her understanding of racism up to the time of the study. The researcher’s interpretation: Though Claire was seen by white community members as engaged in antiracist activities, she could not reasonably be described as being antiracist.
The FIT strategy, in short, revealed that sometimes actions do not speak louder than words. Whether this same insight could have been gained using more conventional approaches to interviewing, we do not know. But because Claire told her own life story and did not simply respond to the questions posed by an interviewer, we can be reasonably certain that the interpretation of the researcher was not merely a reflection of what the researcher expected or wanted to hear. Indeed, the findings about Claire were quite unexpected.
Example #3: When Board Member Participation is Multilayered
Finally, we describe a study in which the FIT was not used but could have been helpful. This example comes from an exploratory study one of us is currently conducting on the relationship between board governance and board members’ developmental levels . We are using a predefined theoretical frame to analyze board participation. Board members are interviewed one-on-one and assessed in terms of their complexity of thought. Then, board meetings are observed, and members’ behaviors recorded. Each individual member-level contribution is coded according to a code scheme developed through a literature review. The final round of data analysis involves comparing individual-level contributions at board meetings to indicators of complexity of mind to identify how, if at all, complexity of mind influences contributions to the board.
This study is theoretically grounded and an example of a study with high information power (Malterud et al., 2016). However, humans are messy creatures. The member-level contributions at the board meetings are reflective of the members complexity of mind but, at the same time, also reflect other attributes such as their tenure on the board, familiarity with the organization, and professional stage of life. A FIT would have helped flesh-out some of these attributes and would have led to a more comprehensive and nuanced analysis of individual-level participation. Specifically, we could have used a FIT to identify (among other questions) how and why the board members began serving, their confidence in their own skills as a board member, and what experiences from their personal and professional life shaped their governance decisions. The answers to these questions would provide important context through which to analyze their contributions to the board.
Caveats
All research strategies have limitations. Consequently, before concluding, we need to add five caveats related to the FIT’s limitations.
First, we have already noted the FIT strategy is most appropriate for documenting changes over time. If a researcher is attempting to explicate research participants’ perspectives and concludes that it is appropriate to treat these perspectives as being relatively static, it is possible that another interview strategy, for example, an adaptation of Spradley’s (1979/2016) interview techniques discussed above, may be more appropriate than using the FIT.
Second, the FIT strategy may produce uneven data across study participants. Some interviewees’ timelines—and their comments about their timelines—are likely to be richly detailed, while other timelines and commentaries may be quite sparse. We have even had research participants who did not want to draw a timeline. In these cases, we let these individuals simply talk through their timeline rather than write it out first.
Of course, variations in the extent of information provided by different interviewees are more a problem with interviewing, in general, than with FIT interviews in particular. But this problem does point up the need, at times, for considerable interviewer probing and, probing, in turn, suggests that researchers may have more impact on FIT data than we have suggested previously.
The third caveat also is a problem FIT interviews share with other types of qualitative interviews: The data generated are self-report data, and interviewees are unlikely to have a complete understanding of the various influences on the organizations and/or programs in which they participate or even on their own lives. Indeed, the information they share will undoubtedly be filtered by their cultural backgrounds, economic class, and education. This problem can be mitigated, somewhat, by conducting FITs with as diverse a sample of participants as possible and triangulating insights from one FIT interview with the results of many other interviews. We also recommend using other research methods, whenever possible. Both of us are fans of mixed-methods studies, for example. Of course, as Mathison (1988) has noted, efforts to triangulate findings will not necessarily produce consistency. They should, however, provide a deeper and more nuanced understanding of the limits of any single FIT interview.
The fourth caveat involves possible interviewee reactivity: The FIT conceivably might influence how participants respond to questions asked in post-FIT parts of the interview process. If participants are primed, during a FIT, to think about formative influences on their philanthropy, for example, they may, consciously or unconsciously, focus their answers to subsequent questions about their giving on giving that appears to be associated with the identified formative influences and ignore other examples of their giving behavior. There is a bit of irony here: a method designed to minimize bias could conceivably introduce bias into a study. Even though we have not experienced this phenomenon when using the FIT, reactivity is certainly possible.
Finally, especially when FIT interviews are used to surface information about people’s personal development rather than the development of nonprofit organizations or programs, ethical issues might arise. The FIT interview with the philanthropist Joseph alluded to above, for example, led to revelations about Joseph’s personal history that Joseph may not want revealed publicly, even if a pseudonym is used. At the very least, the potential for ethical problems to arise when employing the FIT suggests the importance of member checking and giving interview participants the right to have a rebuttal included in research reports. One could even argue that interviewees should have the right to veto details from their lives if one believes ethical concerns about privacy should trump scientific norms of full disclosure.
Concluding Comments
Whether researchers are positivists, constructivists, or representatives of some sort of creative synthesis of these supposedly antithetical research paradigms, they, almost certainly, will want to understand and, presumably, report their research participants’ thinking as accurately as possible. To do this, they must employ interview procedures that allow interviewees to direct and shape the interview process, while still ensuring interviews are conducted efficiently so interviewees’ time is not wasted. Here we have touted the FIT as an interview strategy that can accomplish both purposes. We also have provided examples of how the FIT strategy has been used—or, in one case, could have been used—in nonprofit/philanthropy studies.
We encourage qualitative researchers to incorporate the FIT into their research whenever it is appropriate to do so. We look forward to reading about the insights generated by the FIT on the pages of this and other nonprofit/philanthropy journals in the not-too-distant future.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
