Engendering Research Ethics in Political Science and IR: A Feminist Ethics of Care in the Profession?

Gendered patterns in knowledge production in political science and IR have been widely documented in nearly all aspects of the academic endeavor. Women are underrepresented in publishing in leading journals (Dion and Mitchell Reference Dion and Mitchell2019; Teele and Thelen Reference Teele and Thelen2017), books (Samuels and Teele Reference Samuels and Teele2021), citations counts (Maliniak, Powers, and Walter Reference Maliniak, Powers and Walter2013), and senior ranks of the profession and professional organizations (Alter et al. Reference Alter, Clipperton, Schraudenbach and Rozier2020; Monroe and Chiu Reference Monroe and Chiu2010).Footnote ² The underrepresentation of women scholars also plays out in syllabi (Hardt et al. Reference Hardt, Kim, Smith and Meister2019; Phull, Ciflikli, and Meibauer Reference Phull, Ciflikli and Meibauer2019), while students also undercite work by women (Liu, Devine, and Gauder Reference Liu, Devine and Gauder2020). The practice of research ethics is a part of knowledge production in political science and IR, and is likely to reflect these gender-based differences with implications both for women researchers as creators of knowledge and the nature of the knowledge created.

Publication and citation gaps matter because they can privilege some substantive questions and methods over others (often quantitative methods in political science; see Teele and Thelen Reference Teele and Thelen2017). Women are socialized within these inequalities from the start of their academic journey, with women receiving less quantitative training than men while PhD students (Gatto et al. Reference Gatto, Gohdes, Traber and van der Velden2020). Gender gaps persist because they are reproduced through the “Matilda effect” where “women’s research is viewed as less important or their ideas are attributed to male scholars, even as a field becomes more diverse” (Dion and Mitchell Reference Dion and Mitchell2019, 312; see also Brown et al. Reference Brown, Horiuchi, Htun and Samuels2020; Key and Sumner Reference Key and Sumner2019). These gendered ways in how knowledge is produced and evaluated point to women’s “epistemic marginalization” and “epistemic discrimination” in the profession (Dalmiya and Alcoff Reference Dalmiya, Alcoff, Alcoff and Potter1993; Fricker Reference Fricker2007; Giladi and McMillan Reference Giladi and McMillan2022). We argue that emerging practices of ethics reporting are likely marked by inequalities not only in terms of who reports and what they report on, but also in terms of how such reporting is received.

A feminist ethics of care directs attention to identifying inequalities that adversely affect women compared to men. In the practice of care, this disparity manifests as women’s overrepresentation in caring roles and caring professions. This point has been noted both by scholars of feminist ethics of care and of prosocial behavior, interested in the study of behaviors beneficial to others, such as care (Diekman and Clark Reference Diekman, Clark, Schroeder and Graziano2015).Footnote ³ The prosocial behavior perspective further clarifies the differences in how men and women practice care: women are disproportionately involved in care, conceived in terms of relational commitments to the well-being of the marginalized and disempowered, in contrast to men whose help is conceived as a chivalrous response to those at risk (Diekman and Clark Reference Diekman, Clark, Schroeder and Graziano2015; Eagly Reference Eagly1987; Nielson, Padilla-Walker, and Holmes Reference Nielson, Padilla-Walker and Holmes2017). Gendered prosociality, which highlights women’s orientation to others as a feature of their communal behavior in contrast to men’s agentic behavior defined by assertiveness and self-focus (Farrell and Finkelstein Reference Farrell and Finkelstein2007; Kidder Reference Kidder2002),Footnote ⁴ is triggered by different mechanisms. Environments are one such mechanism “which encourage[s] specific traits or behaviors” (Diekman and Clark Reference Diekman, Clark, Schroeder and Graziano2015, 380).

With the ongoing institutionalization of research ethics in political science and IR, the profession is a type of environment with an expectation for all involved to care. Yet there can be subtle gender-based differences in enacting care as a form of prosocial behavior within the same workplace. Women engage more in relational workplace behavior than men, both vis-à-vis colleagues and vis-à-vis outsiders they interact with in their occupational roles, even though they discharge their roles equally in terms of providing expert advice (Eagly Reference Eagly2009, 648–9; Farrell and Finkelstein Reference Farrell and Finkelstein2007; Lovell et al. Reference Lovell, Kahn, Anton, Davidson, Dowling, Post and Mason1999). Furthermore, women will often voluntarily go beyond their required duties more than men and take on additional tasks and roles which aligns with a relational commitment to others (Kark and Waismel-Manor Reference Kark and Waismel-Manor2005). We expect to observe more women than men demonstrating ethical conduct of research through ethics reporting, and this effect to play out in mixed-gender co-authoring teams.

A feminist ethics of care perspective also requires us to interrogate whether and how research conducted by women scholars, which has been documented as different from men’s, contributes to women’s articulation of different ethical concerns. The articulation of feminist ethics of care stems from appreciating women’s “way of being” in the world, which, in turn, informs their moral standpoint (i.e., a caring standpoint) and their care practices (Gilligan Reference Gilligan1982; Noddings Reference Noddings1984; Ruddick Reference Ruddick1989). Hence, women’s view of research ethics will be directly related to the kind of research they conduct, for example, by topic or data collection. There is a wide gender disparity in different subfields, topics, and regions men and women study in political science and IR (De Haan Reference De Haan2020; Reid and Curry Reference Reid and Curry2019). For example, men scholars are about twice as likely to study “conflict” as “peace,” while women authors are slightly more likely to write about “peace” than “war” (Bright and Gledhill Reference Bright and Gledhill2018).

Likewise, women’s experience of fieldwork differs from men’s. Specific challenges experienced by a lone female researcher in a volatile research site will reflect in their considerations of research ethics (Krause Reference Krause2021). Women are more likely to face harassment—predominantly sexual harassment and other forms of sexual violence—in the context of ethnographic research (Hanson Reference Hanson2019). But men “are also encouraged to endure physical and emotional violence associated with expectations of hegemonic masculinity” (Hanson Reference Hanson2019, 2). Following Gilligan (Reference Gilligan1982), women’s “different voice” can manifest in articulating different ethical concerns from men, including in their openness to discuss the emotional challenges of research (Shesterinina Reference Shesterinina2019). These insights raise the prospects of ethical challenges and dilemmas which women and men may approach and articulate differently, or prefer to avoid.

At the same time, scholars have warned against eliding differences among women and among men. While gender is a “privileged analytic,” scholars must also be attuned to multiple and intersecting axes of diversity (Raghuram Reference Raghuram, Mahmoudi, Brysk and Seaman2021). Beyond challenging the overarching notions of feminine versus masculine care as necessarily distinct and oppositional (Jordan Reference Jordan2019), the intersectional lens brings into the fold multiple identities and their role in structuring gendered practices in performing care (Hankivsky Reference Hankivsky2014; Tronto Reference Tronto2010). They are evidenced elsewhere in the academic profession. For example, women of color are more marginalized in senior academic and leading administrative posts than white women in relation to men. Moreover, women of color have not achieved gains in their representation compared to some improvement recorded for men of color (Silbert, Punty, and Ghoniem Reference Silbert, Punty and Ghoniem2022). Such intersections reveal new planes of domination and inequality, finely texturing our understanding of how care is gendered. Hence, the intersections of ascriptive and non-ascriptive identity axes—including ethnicity, race, seniority (vs. precarity), wealth, privilege, education, institutional prestige, methodology, and methodological training—might constitute complex hierarchies in which some women conceive of and practice the ethics of care differently from other women and men when conducting political science and IR research.

In testing this hypothesis, we follow existing literature that has demonstrated an intersection of gender and methodology to evaluate the effect of this intersection on the reporting of ethics. We also examine other intersections, namely authors’ seniority, and geography of an author’s home institution.

Data and Methods

To identify articles reporting on ethics within these journals, we aimed to locate those that explicitly reported ethics. We searched for articles that contained: “ethics of research,” “research ethics,” “IRB,” “institutional review,” “ethics review,” “ethics procedure,” “ethics committee,” “ethics clearance,” “ethics board,” “human subjects review,” “human subjects committee,” “review board,” “consent,” “anonymous,” “confidential,” “deceit,” “deception,” and “debrief” (and elaborated in Section 1 of the Supplementary Material).Footnote ⁷ We trialed several versions of these search terms and their associated wildcards, including terms such as consent and anonymity, rather than only including references to institutional ethics (e.g., IRB).Footnote ⁸

We recoded Teele and Thelen’s (Reference Teele and Thelen2017) coding of article method to encompass quantitative, qualitative, and mixed-methods articles that use and analyze data.Footnote ⁹ We also code for, and include, conceptual/review articles that reflect on authors’ prior empirical work or discuss methods.Footnote ¹⁰ We excluded articles coded by Teele and Thelen (Reference Teele and Thelen2017) as “political theory” or “formal theory” because we are interested in articles that are empirical and, hence, could be expected to report or discuss research ethics.

In this ethics reporting versus non-reporting dataset, there are 5,629 articles published in nine journals (2000–18). The overwhelming majority (5,392, 95.8%) did not report ethics compared to 237 articles (4.2%) that did (Figure 1). As a percentage, ethics reporting has increased since 2013 (Figure 2)—with few articles reporting ethics published before 2009—but the number of articles published has also increased.

Figure 1. Ethics Reporting versus Non-Reporting

Figure 2. Percentage of Articles Reporting Ethics over Time

To test H1 and H3, we use this dataset to examine differences in ethics reporting versus non-reporting by author gender and article method (quantitative, qualitative, mixed-methods, and conceptual/review). To infer author gender and update this dataset (2015–18), we used the genderizeR R-package that predicts gender based on first names (Wais Reference Wais2006).Footnote ¹¹ We also acknowledge the challenges of measuring gender according to a binary classification of men/women, that excludes trans and non-binary scholars. At the same time, this analysis follows existing scholars’ coding and is a necessary first step in interrogating whether and how there are gender inequalities in ethics reporting.

First, we report descriptive statistics, namely counts of ethics reporting versus non-reporting by author gender and article method. Second, we use binomial logistic regression to analyze the relationship between independent variables (author gender and article method) and the dependent variable (reporting vs. non-reporting).Footnote ¹² We explore two sets of models with a different configuration of gender as the independent variable:

1. 1. numeric variable: percent women (models 1–3),

2. 2. categorical gender variable: women-only (100% women), men-only (0%), and mixed-gender (1%–99% women; models 4–6).

Descriptive Comparison of Reporting versus Non-Reporting by Gender and Method

Remarkably few articles (4.2%) across our data report ethics (2000–18; Figure 1). This figure is low, for example, in relation not only to all articles published but specifically to articles that would be expected to report some aspect of ethics. Based on the analysis of this dataset, we estimate that 23% of articles use data and methods that directly engage with human participants and, therefore, should be reporting ethics (e.g., field experiment, survey, and interviews). We also estimate a further 23.2% of articles that, while not directly engaging with human participants, are using data for which ethical questions might arise, for example, archival, social media, or textual data (see Section 5 of the Supplementary Material). Based on these estimates, about 10% of articles that should be reporting research ethics are actually reporting it (2000–18). These findings indicate that, absent the mandate to report on research ethics in political science and IR publications, non-reporting will by and large be a default practice of all but a small minority.

Analyzing, from a gender perspective, the minority that do report on research ethics reveals multiple inequalities. Most articles published were authored by men-only (62.1%; Table 1), compared to women-only authors (16.3%) and mixed-gender coauthors (21.6%). However, a higher proportion of mixed-gender coauthors (6.1%) and women-only authors report on ethics (5.4%) than men-only authors (3.2%; Table 1).

Table 1. Author Gender by Article in the Ethics Reporting versus Non-Reporting Dataset

Most articles in the reporting versus non-reporting dataset are quantitative (76.8%), followed by qualitative (15.3%), conceptual/review (5.5%), and mixed-methods (2.4%; Table 2).Footnote ¹³ By methods, more mixed-methods articles report ethics (16.1%; Table 2), than others (quantitative: 4.3%; conceptual/review: 4.2%; qualitative articles reporting ethics: 1.7%).

Table 2. Article Method in the Ethics Reporting versus Non-Reporting Dataset

Intersecting gender and methods (Figure 3), these trends are amplified. Mixed-gender coauthors and women dominate reporting in quantitative (6.1% and 5.2%, respectively, vs. men: 3.5%) and qualitative articles (3.4% and 3.4%, respectively, vs. men: 0.6%). Meanwhile, women dominate reporting in conceptual/review (8.8%, mixed-gender: 4.9%, men: 2.9%) and mixed-method articles (18.9%, mixed-gender: 18.2%, men: 14.1%).

Figure 3. Reporting of Ethics by Author Gender and Method

These unequal trends, which point to women doing the ethics reporting “work” that we observe descriptively, stand to be tested statistically.

Statistical Comparison of Reporting versus Non-Reporting by Gender and Method

To test H1 and H3, we need to determine if these differences in reporting by gender and method are statistically significant. We use binomial logistic regression to analyze the relationship between independent variables (author gender and article method) and the dichotomous dependent variable: reporting versus non-reporting of ethics in political science and IR journals. We use the year of publication and the number of authors as control variables.Footnote ¹⁴

The full results of the regression analysis are available in Table 3 and contain models both for percent women (models 1–3) and the categorical gender variable (models 4–6). We include models with only methods variables and controls (journal year, author number, and models 2 and 5), and only gender variables and controls (models 3 and 6; Table 3). We use quantitative articles as the baseline for comparison with other methods (conceptual/review, mixed-methods, and qualitative), and men-only authors as the baseline for the categorical gender comparison (with mixed-gender and women-only authors).

Table 3. Binomial Regression of Reporting versus Non-Reporting of Research Ethics

Note: Odds ratios and 95% confidence intervals in brackets; dependent variable is reporting versus non-reporting of ethics; baseline for models 4–6 is men-only (for gender); baseline for models 1, 2, 4, and 5 is quantitative (for method); the number of observations is 5,629 for all models; models 1 and 4 are the basis for Figure 4. ***p < 0.001, **p < 0.01, *p < 0.05.

Figure 4. Effect of Gender and Method on Reporting versus Non-Reporting of Ethics

Note: Odd ratios and significance values plots of binomial regression models; n = 5,629; Models 1 and 4 in Table 3; see Table 3 for full results. ***p < 0.001, **p < 0.01, *p < 0.05.

We include two gender variables partly for ease of interpretation and usefulness. In model 1, controlling for all other variables, the odds of reporting ethics are multiplied by a factor of 1.006 for every percent increase in women in a coauthoring team ( $ p<0.01 $ )—that is, they are 0.6% higher. As it is not that useful, practically, to talk in terms of increasing the percent of women, we include a categorical gender variable for gender (replicating the coding used in the extant literature). For women-only authors, the odds of reporting ethics are multiplied by a factor of 1.718 (model 4, $ p<0.01 $ ), compared to men-only authors controlling for all other variables—that is, they are 71.8% higher. For mixed-gender coauthors (1%–99% women), the odds of reporting ethics are multiplied by a factor of 1.504 ( $ p<0.05 $ ), compared to men-only authors controlling for all other variables—that is, they are 50.4% higher. These results point to the gendered-nature of ethics reporting, with women leading the practice.

Focusing further on models 1 and 4, both models also indicate a significant relationship between specific methods and ethics reporting (Figure 4). Namely, for mixed-methods articles, the odds of reporting ethics are 3.348 (model 1, $ p<0.001 $ ) and 3.373 (model 4, $ p<0.001) $ times the odds for quantitative articles controlling for all other variables—that is, they are 234.8% and 237.3% higher. The odds of ethics reporting in qualitative and conceptual/review articles are not significantly higher or lower compared to the odds for quantitative articles.

The control variables also demonstrate the importance of other mediating factors. In models 1 and 4, increasing the publication year by one year, the odds of reporting ethics are multiplied by a factor of 1.201 ( $ p<0.001 $ ) controlling for all other variables—that is, they are 20.1% higher. This finding supports our descriptive account of an ethics turn in political science and IR (Figure 2). The number of authors matters, too. Every additional author multiplies the odds of ethics reporting by 1.181 in model 1 ( $ p<0.01 $ ) controlling for all other variables—that is, increases them by 18.1%; however, this effect is not significant in model 4. It might be, for example, that projects with more coauthors are more methodologically and ethically complex or that projects with more coauthors might include members that drive reporting of ethics.

Overall, we find evidence to support H1: that ethics reporting does increase among women authors compared to men, including but weaker in mixed-gender coauthoring teams, controlling for journal year, author number, and research method. The latter being prominent in evidence of support for H3: ethics reporting is not only an issue of gender but intersects also with method.

Data and Methods

To identify relevant articles reporting research ethics, we used the same search terms as in the reporting versus non-reporting dataset.Footnote ¹⁵ We expanded our search to 26 prominent political science and IR journals, broadly defined, in the same period (2000-18) and identified 709 articles published in them that report research ethics. We followed previous scholars’ selection of journals (e.g., Teele and Thelen Reference Teele and Thelen2017).Footnote ¹⁶ We then added other prominent political science and IR journals which also have a broader interdisciplinary reach for authors and readers.Footnote ¹⁷

In determining who is reporting and how ethics is reported in journal articles, and testing the hypotheses related to gender-based inequalities (H1–H3), as in the previous dataset, our focus is on empirical (quantitative, qualitative, and mixed-methods) and conceptual/review articles. We scrutinize published articles as they serve a professional and pedagogical function in setting standards and providing an opportunity for learning, for example, from others’ ethical dilemmas and challenges, as well as demonstrating the ethical integrity of research to wider society. Preregistration plans are likely another source for ethical reporting but have a smaller readership.

Findings: Research Ethics Reporting

We observe an ethics turn in the profession that aligns with its adoption of an ethics of care, with 51.5% of articles reporting ethics published from 2015–18 (Figure 5). Without an established or consistent practice (yet) of reporting ethics, demonstrated by the few articles reporting on ethics (especially prior to 2015), those articles that do report ethics are themselves establishing this practice. As such, they are implicated in knowledge production, particularly in conceiving “good care” in the subfield of research ethics as it pertains to political science and IR.

Figure 5. The Reporting of Ethics Dataset over Time

Reporting of ethics has taken place in specific journals, such as PS (17.2%), JoP (11.8%), and JPART (8.5%)—together 37.5% of articles—compared to other prominent journals in the field.Footnote ²² Ethics reporting also takes place in different locations of articles. For example, most articles reported ethics in the body of the article (54.6%), or in footnotes (31.9%), with fewer reporting in the acknowledgments (16.1%), or in an appendix (6.3%).Footnote ²³ But an important caveat to the ethics turn is the current varied extent of ethics reporting. For example, in most articles, we observed only a short statement about research ethics. Most authors reported either the procedure of formal ethics review (e.g., referencing the permission of the institutional body conducting the research ethics review) and/or procedures of confidentiality and anonymity (e.g., a single sentence); far fewer articles expanded on the specifics of ethical concerns within the context of their research.

By gender, most articles were authored by men-only (44.4%), followed by women-only (27.9%) and mixed-gender authors (27.6%; Table 4). We might expect such a dominance of men given that men dominate publishing of articles overall, as demonstrated by the reporting versus non-reporting dataset. We now examine how this gender composition breaks down by method and the region of a first author’s institution.

Table 4. Author Gender in the Reporting of Ethics Dataset

Most articles that report ethics are either quantitative (56.3%) or qualitative (20.9%), with fewer articles being mixed-methods (9.4%). More articles reporting ethics written by women are qualitative (35.4%) whereas for men and mixed-gender authors most articles are quantitative (64.4% and 69.9%, respectively; Figure 6). We cannot infer whether these findings reflect trends within reporting of ethics, dominant approaches within political science, or publishing biases within journals that favor statistical over qualitative and mixed-methods work.

Figure 6. Articles Reporting Ethics by Methods of Analysis and by Gender

Most first authors in articles reporting ethics are based in North America (71.5%) or Europe/Central Asia (22.3%), compared to other regions. Here, there are not large differences by author gender (Figure 7). Although the practice of reporting ethics is still emerging, more robust institutional procedures of ethics, for example, in North America (e.g., IRB), might spill over into the practice of reporting (in the journals that we analyze). Alternatively, this effect may result from publishing biases within the journals we select that favor North American-based (or trained) authors. Regardless of the mechanism, it appears to affect men and women equally.

Figure 7. Articles Reporting Ethics by Author Gender and First-Author’s Region of Institution

Note: There were no first authors from Sub-Saharan Africa or South Asia.

Finally, in terms of seniority of those who report ethics, first authors are most commonly those in tenure-track (assistant professors: 33.9%), or tenured positions (professors: 24.0%; associate professors: 19.7%), with only a minority who are PhD students (9.9%) or post-docs (5.5%; Supplementary Figure S4).Footnote ²⁴ By gender, in terms of first-author rank, we observe a similar pattern where assistant professors are the largest category for women-only (41.4%) and men-only articles (33.7%), but many articles are also first-authored by men professors (25.1%). For mixed-gender coauthoring teams, first authors are more evenly split across the three categories (Figure S5 in the Supplementary Material). By region, most assistant professors (80.4%), associate professors (71.4%), professors (71.2%), PhD students (77.1%), and other students (e.g., BA/MA) are based in North America, which also speaks to the North American dominance of ethics reporting (Figure S7 in the Supplementary Material). Without access to wider data on publishing across ranks (which to the best of our knowledge does not exist), we can only infer that these trends may reflect broader publishing patterns and gendered-patterns across academic positions. We could infer, for example, that with most articles that report ethics being authored by pre-tenured academics, that is, assistant professors, there may be a trend toward greater training or socialization in research ethics in North America,Footnote ²⁵ which further highlights the pedagogical function of ethics reporting.

From this descriptive analysis, we can build on the findings from the previous dataset which suggest that women and mixed-gender coauthors are reporting ethics more than men, once accounting for the number of articles published (H1). We can also begin to see other author attributes that intersect with gender (H3), such as method and seniority, as well as attributes that do not amplify gender differences, such as the region of authors’ institutions.

Dimensions of Ethics by Gender

We now turn to how dimensions of ethics are reported, to capture the epistemic aspect of the emerging practice of ethics of care in the profession. We investigate whether women and men report on different dimensions, and, if so, how (H2). Following our codebook, we examine several dimensions of ethics reporting: formal reporting of ethics, everyday ethics, confidentiality and anonymity, consent, risks, and benefits.Footnote ²⁶

By gender, we find significant differences in reporting confidentiality and anonymity, everyday ethics, risks, and benefits, with women reporting more than men-only and mixed-gender coauthoring teams (Figure 8 and Table 5). For example, 65.2% of women-only authors report confidentiality and anonymity (Figure 8), compared to 54% of men-only and 53.6% of mixed-gender coauthor teams ( $ {\chi}^2 $ ( $ df=2 $ , $ N=709 $ ) = 7.488, $ p=0.024 $ ), and 28.8% of women report risks of research compared to 19% of men-only and 16.8% mixed-gender coauthor teams ( $ {\chi}^2 $ ( $ df=2,N=709 $ ] = 9.945, $ p=0.007 $ ; Table 5). Although we see men slightly more reporting formal ethics than other groups (men-only: 56.8%; mixed: 52%; women-only: 51.5%), this difference is not significant. Differences by gender in reporting consent are also not significant.

Table 5. Chi-Squared Tests for Differences in Proportions by Gender and Method

Note: For tests on method, conceptual/review were removed to avoid overdetermining results by nonempirical papers. Everyday ethics and benefits are not reported for methods because of too few observations.

Figure 8. Reporting of Ethics by Dimension and Author Gender

Despite the small number of articles being compared here, especially when broken down by gender, this analysis has shown via granular coding of ethics reporting that the dimensions of ethics that are reported relate to gender in meaningful ways. Women give more attention to reporting confidentiality and anonymity, everyday ethics, risks, and benefits versus men-only and mixed-gender coauthoring teams (H2, though not where women coauthor in mixed-gender teams). Moreover, there are no dimensions of ethics that men-only and mixed-gender coauthors are reporting more than women.

Dimensions of Ethics and Method

To return to H3, we now analyze if and how reporting of ethics relates to method (we exclude conceptual/review papers from our analysis here and from χ² tests to avoid nonempirical papers overdetermining the results).

We find significant differences in reporting formal ethics, consent, and confidentiality and anonymity by method (Figure 9 and Table 5). In reporting formal ethics, quantitative articles report far more (65.9%) than other methods (21.6%–26.9%; $ {\chi}^2 $ ( $ df=2,N=614 $ ] = 102.25, p<0.001). Consent is also more likely reported by quantitative articles (29.1%; Figure 9) compared to mixed-methods (19.4%) and qualitative (17.6%, $ {\chi}^2 $ ( $ df=2,N=614 $ ] = 8.932, $ p=0.011 $ ; Table 5). However, qualitative (88.5%) and mixed-methods (85.1%) are more likely to report confidentiality and anonymity than quantitative articles (43.6%; $ {\chi}^2 $ ( $ df=2,N=614 $ ] = 111.16, p<0.001). By contrast, differences of reporting of risks by method are not significant.

Figure 9. Reporting of Ethics by Dimension of Research Ethics and Article Methodology

We also see differences in which dimensions are significant for gender but not for method, and vice versa. For example, formal ethics and consent are not significantly different by gender but are by method (Table 5).

We, therefore, find further evidence for H3: that how ethics is reported is not only explained by gender but also by method. These differences are notable since formal ethics, consent, and confidentiality ought to be similarly important regardless of method used. But, as we find, such similar ethical concerns do not make equally their way into the reporting of ethics in political science and IR.