The focus for January is on researchers’ roles, including characteristics and skills critical to success. Read the whole series here.
Breaking from tradition is necessary for the creation of new knowledge. Copernicus, Newton, Lavoisier, and Einstein are all known for shattering the scientific tradition of their age. Funding agencies recognize this, and aim to support risky, path-breaking projects: National Institutes of Health, through the High-Risk, High-Reward Research Program; the National Science Foundation through its Early-concept Grants for Exploratory Research (EAGER); and the European Research Council through Horizon 2020’s support for “unconventional and innovative approaches.”
Although breaking from tradition may advance science, it may not benefit scientists themselves. This is because of the inherent valuation risk: the risk that scientists who break from tradition may be evaluated as incompetent or scattered. Peers may assume they are incompetent and incapable of doing the conventional thing and thus must resort to something different, unusual, and of presumed lower value. Or, they may assume they are a scattered dilettante, without a commitment to a clear professional identity. In science, these perceptions come to the fore during the peer review process (when reviewers and editors decide whether one’s work is good enough for publication) and in university hiring and promotion processes (when committee members decide whether to hire or promote a scientist).
Given this, one might ask, why researchers go to the effort of undertaking unconventional, path-breaking work? What makes some scientists more likely to engage in research that breaks from tradition, despite the risks? In our recent study, we considered two possible explanations. First, we thought that scholars affiliated with high-status demographic groups (like men), and high-ranking organisations (like scholars at prestigious universities) would be less susceptible to valuation risk, and better able to dismiss or overcome its possible ramifications, and thus gravitate toward unconventional methods. Second, we thought that scholars who have already committed to an academic identity that is consistent with the use of unconventional methods will be more likely to use them. This is because having a consistent research identity is rewarded professionally, and helps one bear the valuation risks associated with pursuing unconventional methods.
The identity of interest in our research was that of an interdisciplinary scholar. Identifying as an interdisciplinary scholar (one who is open to ideas and theories from other disciplines) is consistent with (and likely prompts the use of) unconventional methods, especially as many conventional social science methods are imported, wholesale, or in their key machinery, from other disciplines (e.g., correlational analysis, multiple regression, experimental design, and hypothesis testing were initially developed to study crops, fertilizers, and genetics).
To test these ideas, we collected and analysed quantitative and qualitative data about the use of three unconventional methods in sociology: Correspondence Analysis; Qualitative Comparative Analysis (QCA); and Sequence Analysis. The quantitative data were derived from peer-reviewed articles that use one of these methods, and a comparison group of articles on similar topics that do not use one of these methods. The qualitative data were derived from interviews we conducted with published, unpublished, and likely users of these unconventional methods. We asked interviewees specifically about how they manage the career risks associated with unconventional method use.
In general, we found that scholars who are male or affiliated with top-tier universities, as well as those already committed to an interdisciplinary identity consistent with the use of unconventional methods, are more likely to use them in their published work. But we also find that these associations depend on how unconventional the method seems. QCA is the most unconventional method, as it presents an approach to causality that is unique in sociology, and its connections to traditions (in other fields) are more obscured. Members of high status groups are less likely to use QCA in published work and an interdisciplinary identity actually decreases the likelihood of use.
From the interview data, we identified five successful strategies scholars use to manage their use of unconventional methods. The most common strategies used to manage valuation risk involved demonstrating competence in conventional methods, which likely served to decrease the chance that their use of an unconventional method was evaluated as incompetence. This was accomplished in three main ways: demonstrating conventional expertise in one’s research program; incorporating conventional expertise into the application of the unconventional method (even when doing so ran counter to the logic of the method); and emphasizing the technique’s similarity to conventional methods. The fourth and fifth involved demonstrating commitment to an established identity: affirming one’s identity through method use or distancing one’s identity from method use through disclaimers. Finally, two strategies that were largely described by informants as unsuccessful were fundamentally different from the successful ones in that rather than acknowledging and trying to manage valuation risk, they were attempts to gain status or a positive identity through the use of an unconventional method.
Our own study was also unconventional in a few ways. Whereas most research on unconventionality focuses on ideas, we focused on methodological tools. We were also able to identify early indicators of the use of unconventional methods and highlight the important link between status derived from association with elite research organisations and the ability to limit valuation risk. Although we already know that highly influential science is both traditional and conventional, we drew upon qualitative data to elaborate how scholars strategically incorporate convention into unconventional work in the practice of science.
Our quantitative results suggest that science policymakers aiming to encourage unconventional research should support scholars who are not members of high-status groups and organizations (i.e., NSF ADVANCE program) and cement commitments to unconventional-friendly affiliations early in scientists’ careers (i.e., the Andrew W. Mellon Foundation’s New Directions Fellowships). Our qualitative results, however, provide an important caveat. We find that scholars who use unconventional methods successfully (i.e., publish it in a journal) also tend to demonstrate competence in conventional methods in a variety of ways. These individual-level strategies that are largely absent from current discussions about how to promote “high risk, high reward” research. Though many important scientific problems require cross-disciplinary solutions, scientific careers are still largely organized by disciplines and a scholar who is not a proficient user of the conventional methods in her discipline is unlikely to significantly contribute to that discipline.