WHO ARE THE COLLABORATORS?

With whom should we collaborate on a given project? Who are the stakeholders? Another way to ask this question is: Who has the power to shape the research we do? One potential answer is that anyone within a particular community can be a stakeholder. Alternatively, in a more purposive approach to stakeholder selection, a researcher might turn to a select group of individuals and ask for their involvement. These individuals could be leaders in a social movement or teachers at a school—gatekeepers, in other words, to the community of interest. Collaborators also might emerge in a sort of “snowball” fashion: the researcher’s first contact then directs her to additional contacts, who help her to identify a few more.

I list three of potentially myriad selection methods here, each of which, I think, has weaknesses. Gatekeepers, for example, are power holders. Whereas they may be legitimate (in the Weberian, arrived-at-their-position-legally sense of the term) leaders, gatekeepers may not be representative of the community at large. They may not even be popular. Alternatively, our first contact within a community could be a product of luck or status. Identifying additional stakeholders from there might produce only one specific view about a question or phenomenon.

Conversely, can anyone be a stakeholder? Although any community member, in theory, can have a stake in research about their community, structural obstacles and power hierarchies may get in the way. For example, do community members have the freedom—of time, of status, of obligations—to provide their voice? Do they have access to the project? Furthermore, do they even know the project exists?

I do not think stakeholders should be a “representative sample” of the broader community; neither do I think a single-selection logic should drive the choice of stakeholders for all CM-based projects. Yet, one benefit of CM is that it helps to derive more valid measurements of key concepts. Indeed, CM practitioners take great care to demonstrate that the collaborative data-collection process achieves this methodological goal. For example, the researchers behind the Everyday Peace Indicators projects have implemented a multi-step approach to first solicit, then craft, and then verify the set of indicators used to describe everyday peace in a community (see https://everydaypeaceindicators.org). Practitioners also have analyzed whether these bottom-up indicators broadly match the more conventional (i.e., top-down) indicators used by existing and widely used databases (Firchow and MacGinty Reference Firchow and MacGinty2017).

Stakeholder selection is vital to the objective of greater measurement validity. The logic of selection will shape a researcher’s understanding of the meaning that a community attaches to a concept because it defines the researcher’s primary interlocutors. Who we include as stakeholders, ultimately, matters greatly.

In considering the logic of selection for our collaborators, I make two points. First, because collaboration cannot involve all community members, the type of selection process used to engage with some of them should be scrutinized. Indeed, no method escapes addressing this type of problem. With focus groups, for example, researchers typically articulate a set of requisites that potential individuals must meet to qualify as participants (Cyr Reference Cyr2019). These selection criteria are derived from the questions that researchers seek to answer. Importantly, the cogency of the project’s overall findings can be assessed as a function of those clearly specified requisites. Practitioners of CM face this same standard.

My second point is an extension of the first. In devising a set of criteria for identifying stakeholders, a collaborative researcher is necessarily establishing a boundary between who can and cannot participate in the collaborative process. The methodological necessity of selection, in other words, may run counter to the normative goal of CM, which is to privilege community voices in the research process. We can imagine a situation in which a person in a community of interest approaches a researcher. This person does not fit the criteria established for defining stakeholders but nevertheless wants to participate in the collaborative process. What does the researcher do in this instance?

The immediate answer, I think, is to acknowledge that an “unqualified” (from the point of view of the selection criteria) individual approached the researcher to collaborate; explain the justification for including (or not) this individual; and evaluate the potential impact of their inclusion (or not) for the overall shape of the project and its findings. Nevertheless, the need to purposively select stakeholders hints at a potentially larger concern: Can the methodological and normative goals of CM be met simultaneously?

Another concern regarding who can be a stakeholder is germane to this discussion. CM allows research subjects to be a part of the production process. This agency is a direct response to decades of extraction on the part of the academy. CM is particularly valuable, therefore, when working with marginalized and/or minoritized groups. Gellman’s (Reference Gellmann.d.) work with Yurok-language students is a clear example of this. Before initiating the project, Gellman had to obtain permission from the Yurok Tribal Council. The Council was distrustful of any outsider, given their history of genocide and exploitation by white settlers (see Gellman’s article in this symposium). Ultimately, she gained access to the students only after committing to work with the Council to develop a research project that would serve the Council’s needs as much as hers. A collaborative project is one that addresses stakeholder interests in addition to those of the researcher. Both groups, as a result, gain from the experience.

The value of this type of engagement with historically silenced groups is clear. However, not all groups have experienced this type of structural exclusion. The question then becomes: With whom do we have an ethical obligation to collaborate? Should we collaborate with non-marginalized groups? What about with elites? Do we have an obligation to them as well, or does their historic position of power mean that we can be less concerned about simply taking from rather than collaborating with them?

As I ponder the outer limits of CM as an approach to research, it would seem we have an ethical obligation to collaborate even with historically privileged groups. The act of researching on any individual, without offering something in exchange, is still problematic. Moreover, the power hierarchy within the research relationship still exists: In conventional fieldwork, I, as researcher, take from you, as subject. If the claim is that the extraction of data without some kind of reciprocity or shared experience is ethically problematic, then it should be so regardless of who I am researching.

Perhaps I am taking the question of ethics too far. Collaborative research involves working with groups so that both sides of the research relationship (i.e., the researcher and the subject) can gain from the experience. These gains are achieved most directly when the subject is allowed to participate in the research process. In this sense, collaboration implies ownership, or “responsible agency in the production of knowledge and the improvement of practice” (McTaggart Reference McTaggart1991, 171). Consequently, a different way to think about with whom we must collaborate may arise from the perspective of ownership. Are we ethically obliged to share research ownership with those who are already long-standing “owners” of power (i.e., elites, politicians, large landholders, and white males) in a particular place? Perhaps not.

When we study groups in historical positions of power, the methodological goals of CM may take precedence over ethical concerns. Collaborating with elites on at least certain parts of a research project may yield unique insight about the questions we should ask and the people with whom we should speak. This is especially the case when we want to understand, for example, structures of power or political decision making.Footnote ¹ The goal of greater conceptual and measurement validity, therefore, may be what motivates us to collaborate with, for example, elites—especially for those topics with which they are intimately and uniquely familiar.

As a final point, it may be that the “whither elites” question is largely irrelevant. A strong affinity may exist between the epistemological position of researchers and the types of questions they ask. Qualitative scholars, broadly speaking, ask “causes of effects” questions; quantitative scholars, by contrast, inquire about “effects of causes” (Goertz and Mahoney Reference Goertz and Mahoney2012). By this logic, the types of researchers oriented toward a collaborative methodology may be unlikely to study groups that, historically, have been in positions of power and privilege. The drive to collaborate may emanate, in part, from a desire to learn from and with those who have not written history. If so, then the question of with whom we collaborate in some ways may be moot.

Nevertheless, we cannot know whether the question is moot until we know more about the researchers who choose to collaborate. Overall, as a potential practitioner of this approach to research, some general guidelines on who the stakeholders can and should be is useful.

THE EXTENT OF COLLABORATION

What do we do when, after committing to a collaborative methodology, our identified collaborators choose not to work with us? Or, what happens if we decide we do not want to revise or change questions as a result of the collaboration? Finally, what if supposed stakeholders are not interested in collaborating but have no problem with us pursuing a more “extractivist” approach? What, in other words, is the extent of collaboration that is required of us by this methodology? What are its limits?

These questions address, in part, the issue of control. A collaborative project, by definition, requires that the researcher relinquish some control over how the project unfolds. More collaboration implies less control. Certainly, this can be a good thing. An additional set (or sets) of eyes can open up researchers to possibilities that they had not considered on their own. Collaborators may identify previously ignored variables or interpret data differently (Belgrave and Smith Reference Belgrave and Smith1995, 85). From a methodological standpoint, collaboration can help to ensure that the research yields credible findings.

Yet, ceding control over a project also raises the risk of transforming it in ways that make it, for example, more difficult to carry out or less readily publishable. The latter concern may matter little to the stakeholders, but it can be important to a scholar—especially a junior scholar who needs a job or is working toward tenure. Likewise, we know that collaborative projects can yield more effective policy (see www.everydaypeaceindicators.org). Should researchers lobby for policy implementation once the better policy is designed? Should they actively advocate for change alongside their stakeholders (Scheper-Hughes Reference Scheper-Hughes1995)? The debate about whether scholars also can be activists is ongoing (Hale Reference Hale2008).Footnote ² Merging those roles can be risky for academics, especially junior scholars. One real concern, then, is that the practice of CM will reveal that the researcher and the stakeholders have different or even competing goals regarding joint projects. (Kaplan’s contribution to the symposium addresses this potential dilemma in more detail.)

I also wonder when does collaboration end? At what point does the stakeholder no longer have a stake in the project? Can a stakeholder ask for a final, prepublication draft to be changed? That it not be published? That it be retracted? The growing literature on member-checkingFootnote ³ (Quatrini Reference Quatrini2020) raises these questions without finding easy answers. I do not want to hold CM to a standard that other methods rarely meet. I do think, however, that collaboration is difficult to achieve in practice without placing at least some limits on the exchange. In addition to community stakeholders, researchers must consider funding obligations, issues of timing, and—again—the overarching imperative to publish. These other facets of research may need to be addressed at the expense of stakeholder wishes.

My point is not only that collaboration can work at cross purposes with other research demands. It also is that researchers, when pressed, can limit collaboration when they have to and, as a consequence, ultimately may reinforce the existing power hierarchy between the researcher and the researched. This fact raises serious concerns about the extent to which CM can fully meet its normative goals.

Finally, when we consider a collaborative methodology, we are inherently taking on more work. Gellman (Reference Gellman2021) had to learn an entirely new method to carry out the survey that the Yurok Tribal Council requested. A collaborative project also undoubtedly involves additional work when it comes to Institutional Review Board (IRB) approval of a research design. For example, a researcher may be asked to include two consent processes: one prior to the initial collaboration and another as a part of data collection.Footnote ⁴

This additional work, in principle, is not problematic. However, learning new data-collection methods or tackling a more complex IRB request may disproportionately handicap junior scholars and graduate students, who often face pressure to complete original research while juggling multiple other tasks. At least some of these costs can be assumed by training programs, such as the two-week summer school at the Institute for Qualitative and Multi-Method Research. Nevertheless, undertaking a CM-based project—given the additional time, energy, and investment required—may be riskier for scholars whose academic careers are not yet secured.

Ultimately, it seems to me that collaboration is accomplished most easily early in the project, after researchers have developed a broad set of research queries but before they have drilled down to the specific questions on which they will collect data. At that early stage, a researcher would have less stake, presumably, in the exact direction that a project takes and therefore would be more open to feedback and influence. Collaboration can mean adding more questions to an already established research agenda—but surely it can (and perhaps should?) involve so much more. After all, if a stakeholder balks at a question or method that is central to the research design, would not collaboration imply at least entertaining the possibility of letting that question or method go?

Collaboration means working with other stakeholders and not giving in to every suggestion. Yet, how do we guide a researcher on achieving this balance? When has our ethical responsibility been met? There is no one answer, of course. Nevertheless, these questions encourage us to consider how far a researcher must go to meet the normative expectation(s) of collaboration. On this point, it might be useful for researchers to develop a clear roadmap of how they envision collaboration to occur. They can identify those aspects that must remain as is and other places where collaboration is welcome and even sought out. In other words, researchers can assert control over certain parts of the project before opening up other parts to collaboration.

One conclusion of the questions raised herein is that the dual objectives of CM—that is, the methodological objective of greater conceptual validity and the normative objective of working with a community rather than extracting from it—at times may work at cross purposes. The goal of inclusion may hinder a researcher’s ability to retain control over the project and craft an appropriate mix of stakeholders. The desire for more valid concepts, by contrast, may push researchers to limit the extent of their collaboration with community members.

A solution to this tradeoff may be in clarifying where and how collaboration occurred. We can imagine collaboration occurring on a continuum.Footnote ⁵ At one end of the continuum, there is no collaboration whatsoever. This would be where most conventional research practices in political science currently reside. The other end of the continuum represents projects that are totally collaborative. This latter extreme—that of total collaboration—is an ideal to which researchers strive but that is impossible to achieve in practice. Collaborative researchers would specify in their work where a particular project fell on the continuum. They also would justify the logic behind why collaboration occurred in some areas and not others and explain how collaboration impacted the evolution of the project. As I stated at the outset, we must not hold any method to an unattainable standard. Some collaboration may be better than none at all.

A CALL FOR COLLABORATION, AFTER ALL

I acknowledge that a healthy amount of skepticism underscores my thinking about the possibilities of collaboration. Nevertheless, I think collaboration is probably more important now than ever. For me, having more clarity about CM will make the practice of collaboration easier. Indeed, I am many (so many!) months into a state-imposed, total quarantine in response to the COVID-19 pandemic as I work on this article. Like others, this time in shelter has been difficult for me as a scholar. As the world ostensibly shuts down around me, I have started to question what value I bring to light by studying niche topics and producing work that is rarely read by more than a few people. What, I wonder, is the point?

One answer comes from Cramer’s (Reference Cramer2016, 446–47) call to a return to community-engaged scholarship. She reminds us that “we have something to learn” from listening to people outside of academia. Lived experience is incredibly valuable. When we turn to communities as sources of evidence, we also should ask how we can contribute to their lives—how we can be of service to them. Fieldwork can—and should—be reciprocal. Engagement makes our data collection beneficial to us, as researchers, and also potentially to the places we study.

Engagement, of course, is at the heart of CM, as researchers actively attempt to bring communities into the knowledge-building process. When successful, the rewards—for the community, for the researcher, and for knowledge—are likely to be quite high. Our pursuit of knowledge as scholars should be meaningful to the people we study. This seems obvious and yet it feels like a rare accomplishment. In collaborating more, we may find that we can advance knowledge and also create meaning.