Introduction
Different branches of science are sometimes compared to “games” that are played by slightly different rules1. In interdisciplinary research and multidisciplinary journals, it can be difficult for editors, reviewers, and readers to assess whether scientists have correctly followed the rules because multiple games are played simultaneously. One of the rules that has recently confused the scientific world concerns positionality statements, which refer to disclosing the scientist’s position in relation to the conducted research. While positionality has historically and pragmatically established functions in certain methodological traditions2–4, in the past few years, an increasing number of scientists across various disciplines have started disclosing their positions5–7. This essay aims to clarify positionality as an epistemological concept and research tool to help scientists in all fields decide whether such statements can contribute to their work.
What positionality does
Game theory has historically distinguished between games with perfect and imperfect information8. In the former, like chess, all information is visible to all players. In the latter, such as a poker, information is hidden from some players. A similar spectrum of information is applicable to science. Some methodologies operate with data and analyses that are mostly visible, whereas others do not reach transparency to the same degree. The epistemological treatise of positionality originates mainly from these “games of imperfect information,” such as qualitative research traditions, where both producing and analyzing data depend heavily on the tacit interactions of individual researchers3,4. A statement of the researcher’s position can serve as metadata that sheds light on both data generation and analysis.
Reflecting on the scientist’s position has long been a central part of anthropological and other interpretively oriented research paradigms2,3. Although positionality can be related to biases, it is not a matter of bias per se but a much larger question of calibrating the research process. When the scientist itself is part of the toolset that is used to generate primary research data—for instance, ethnographic fieldnotes—the details of that “tool” are relevant to acknowledge, not unlike recognizing and transparently reporting materials in biology, physics, or psychology. Because engaging in fieldwork is a process where only a small part of interactions and observations can be documented, it matters who engaged and selected the field note data. In these cases, reflecting on the relevant aspects of the scientist is part of the research process and also makes the scientific community, including meta-analysts, better equipped to assess the study and its components.
Compared to quantitative analytic traditions, qualitative research methods often depend heavily on the scientist’s interpretive effort and are thus even more flexible when it comes to researcher degrees of freedom9. Positionality statements can provide contextualizing information that explains some of the differences between two or more interpreters. For example, a non-Korean scientist who manually codes medical interview data from South Korea may be unable to identify signals that are evident to Korean scientists. Meanwhile, the same non-Korean scientist, as an “outsider,” may be positioned to see other meanings more clearly. Reflecting on such positions can serve as a helpful analytic tool, especially in proactive use. As in preregistration more generally10, stating pre-study assumptions can help making useful analytic calibrations before it is too late. In registered reports, positionality statements may also help inform early peer feedback11.
Two common misunderstandings
A common misunderstanding is that scientific positionality statements must pick a set of group identities, as from a ‘shopping list’12. Nonetheless, there are no reasons to expect selected group identities to be relevant for the generation and analysis of all research questions. Each positionality statement should fit the methods and topics of a specific study: What are the scientists’ backgrounds, beliefs, and other aspects that are likely to affect the generation or analysis of these particular data? In a psychiatric interview study, the scientist’s (lack of) clinical background is likely to be relevant, whereas an ethnographic study of gender interactions may benefit from reflecting on the scientist’s gender4. Because it can never be perfectly known what aspects of a position meaningfully affect the data and analyses, pre-study reflections on positionality serve essentially as auxiliary hypotheses: explicit assumptions about how a scientist’s subjectivity can affect upcoming findings.
The second enduring misunderstanding is that positionality rests on standpoint epistemology or postmodernism, which may sometimes even challenge the possibility of scientific knowledge2,6. Reflections of positionality do not demand a specific epistemology, yet epistemological (and ontological) assumptions can belong to a stated position. As the present examples illustrate, positionality statements can contribute to values that align with objectivity and traditional philosophies of science, such as those based on falsification13. Nonetheless, positionality statements should always be used knowingly with a function that serves the study in question. For example, multiauthor studies that operate with numerous experts may benefit more from a summary of expert views than from hundreds of statements. For research designs that deal with thematic political analyses by one or two authors, more detailed statements of positionality can be advantageous. Because positionality does not equally benefit (or harm) all studies, journals should not mandate positionality statements but treat them like other research tools: encourage them to be used wisely when fitting with the philosophy of a study.
To position or not
How should scientists decide when to use a positionality statement? The first step is to reflect on one’s own role within the chosen methodology and topic: the more weight a research design has on individual scientists and their implicit decisions or interactions, the more likely a position is to impact the results, especially if the research question has significant subjective relevance. Although the role of individual scientists, for example, in controlled statistical designs may be minor, the increasing number of adversarial collaborations and unresolved disputes over meaningful effect sizes imply a need for case by case reflection14. Positionality is not a bias, but can turn into one if a scientist fails to acknowledge its true effects.
Second, the relevance of positionality does not differ from that of setting other assumptions for a study. Reflexive self-assessments will sometimes produce false negatives and positives, as do other assumptions, but the very act already contributes to useful critical positioning. Openly reflecting on one’s own role in the research process subjects it to the same scrutiny as other research materials15. In the same way as it is impossible to guarantee that reported data exclusions, manipulations, and other research decisions correspond to reality16, the degree to which positionality statements echo actual positions can never be fully known. Nonetheless, spelling out positionalities and their (lack of) relevance remains one of the tools that can help scientists not fool themselves as long as the tool is used knowingly.
Acknowledgements
As this essay did not involve data or analysis, I do not consider a positionality statement to serve any meaningful function here. However, it may be worth disclosing that I have both used and omitted positionality statements in my previous work. One could argue that such facts belong to the competing interests, as the present essay potentially reframes my earlier methodologies.
Faculty Opinions recommendedReferences
- 1.
Kuhn T:
The Structure of Scientific Revolutions. University of Chicago Press, 1962. Reference Source
- 2.
Salzman PC:
On reflexivity.
Am Anthropol.
2002; 104(3): 805–811. Publisher Full Text
- 3.
Malterud K:
Qualitative research: standards, challenges, and guidelines.
Lancet.
2001; 358(9280): 483–8. PubMed Abstract
| Publisher Full Text
- 4.
Knott E, Rao AH, Summers K, et al.:
Interviews in the social sciences.
Nat Rev Methods Primers.
2022; 2: 73. Publisher Full Text
- 5.
Zamzow R:
Scientists clash over 'positionality statements'.
Science.
2023; 382(6670): 501. PubMed Abstract
| Publisher Full Text
- 6.
Savolainen J, Casey PJ, McBrayer JP, et al.:
Positionality and Its Problems: Questioning the Value of Reflexivity Statements in Research.
Perspect Psychol Sci.
2023; 18(6): 1331–1338. PubMed Abstract
| Publisher Full Text
- 7.
McKerracher L, Núñez-de la Mora A:
More voices are always better: Tackling power differentials in knowledge production and access in human biology.
Am J Hum Biol.
2022; 34(Suppl 1): e23712. PubMed Abstract
| Publisher Full Text
- 8.
Morgenstern O:
Game theory: theoretical aspects. In: International Encyclopedia of the Social Sciences. Macmillan and Free Press, 1968; 62–69.
- 9.
Simmons JP, Nelson LD, Simonsohn U:
False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant.
Psychol Sci.
2011; 22(11): 1359–1366. PubMed Abstract
| Publisher Full Text
- 10.
Hardwicke TE, Wagenmakers EJ:
Reducing bias, increasing transparency and calibrating confidence with preregistration.
Nat Hum Behav.
2023; 7(1): 15–26. PubMed Abstract
| Publisher Full Text
- 11.
Karhulahti VM, Branney P, Siutila M, et al.:
A primer for choosing, designing and evaluating registered reports for qualitative methods [version 2; peer review: 2 approved].
Open Res Eur.
2023; 3: 22. PubMed Abstract
| Publisher Full Text
| Free Full Text
- 12.
Folkes L:
Moving beyond ‘shopping list’ positionality: Using kitchen table reflexivity and in/visible tools to develop reflexive qualitative research.
Qual Res.
2023; 23(5): 1301–1318. Publisher Full Text
- 13.
Popper K:
Conjectures and Refutations: The Growth of Scientific Knowledge. (Routledge), 1963.
- 14.
Corcoran AW, Hohwy J, Friston KJ:
Accelerating scientific progress through Bayesian adversarial collaboration.
Neuron.
2023; 111(22): 3505–3516. PubMed Abstract
| Publisher Full Text
- 15.
Mason J:
Qualitative Researching. (Sage), 2002.
- 16.
Bishop D:
How Scientists Can Stop Fooling Themselves over Statistics.
Nature.
2020; 584(7819): 9. PubMed Abstract
| Publisher Full Text
