What are the injustices of the world? What causes them? How might we mitigate them? Feminism needs empirical inquiry in all subfields to inform our understanding of the world and our normative reflections on it. Feminist theorists do not necessarily need to do it ourselves, but we do need it. Because we need empirical research, we need to be able to evaluate its methodology. In international relations, feminist scholars adopt and develop methods previously developed in other fields for nonfeminist questions. Feminist theorists need to evaluate this adoption and development by feminist IR in order to learn all we can and should from feminist IR scholarship.

Feminists using both qualitative and quantitative methods have reflected that they find their research constrained by debates in the field about positivism and methodology. Conversations between Ann Tickner and Robert Keohane seemed at cross-purposes at best and to close off, not open conversation at worst. A normative feminist theorist asks a positive theorist to reconceive the conceptual building blocks of the IR field (Tickner Reference Tickner1997) and a positive theorist asks a normative theorist for a positive research agenda (Keohane Reference Keohane1998). Although they seemed not to understand each other, feminists in IR felt compelled to work in response to this debate. Moreover, despite a feminist professed openness to a broad range of theoretical approaches, even some researchers perceived positive approaches as unfeminist or incompatible with feminism because they narrow the bands of methodological reflection to questions of hypotheses, method, and analysis. Such debates potentially undermine the important contributions of feminist theory to feminist IR and of feminist IR to feminist theory.

With important exceptions, particularly for the United States (Ackerly and True Reference Ackerly and True2008; Maliniak et al. Reference Maliniak, Oakes, Petereson and Tierney2007), students have had to learn to adopt, develop, and assess research methods as applied to feminist questions on their own. They receive quantitative training in their departments and find qualitative methods and guidance in feminist empirical research outside of IR. The field has developed such that now some departments offer feminist and qualitative methods training. Yet although some are getting this training in their programs, there is still limited published work that explains feminist researchers' methodological dilemmas and how they are resolved.

In the edited volume Feminist Methodologies for International Relations, Maria Stern, Jacqui True, and I sought to create a text in which feminists made visible the methodological reflections that often were not visible in the final publications of feminist IR research (2006; reviewed in this journal by Anna M. Agathangelou and Heather Turcotte, Reference Agathangelou, Turcotte, Ackerly, Stern and True2008). In the middle of that book, Stern, Tami Jacoby, Bina D'Costa, and Carol Cohn detail their methods and methodological reflections. Their frank reflections on the challenges of particular methods, which included oral history, discourse analysis, participant observation, historical methods, and interviews, are important illustrations of the ways that feminist IR scholars use these tools in their research.

Further, two of the authors exhibit a form of methodological reflection that is particularly provocative. Stern changed her theoretical perspective from standpoint to poststructuralist in order to cocreate her informants' narratives of security on which she based her discourse analysis. D'Costa changed her research question from rape as a weapon of national conflict to gendered nation building. For both, the profound shifts in thinking took place “in the field” and were guided by their understandings of feminism as a research ethic—a theoretically informed practice that affects every decision about research, from theoretical predisposition to question to method of data collection to method of analysis to publication. They exhibit feminist rigor in research, a rigor directed at reflection on all stages of the methodology.

This rigor is the bridge between feminist theory and empirical inquiry. When reflecting on our methodological choices, feminist empiricists can use feminist theory. And when feminist empiricists are transparent about the theoretical influences on their methodological choices, feminist theorists can help bridge the gulf created by normative versus positivist and other methodological debates.

Laura Parisi, Clair Apodaca, and Mary Caprioli exhibit such rigor in their reflection on their own methodological challenges throughout the research process. In “The Numbers Do(n't) Always Add Up: Dilemmas in Using Quantitative Research Methods in Feminist IR Scholarship” Parisi reflects on all stages of her quantitative methodology, including the possible political implications of her findings and their interpretation (this issue, pp. 410–419). She decides that such considerations should inform how she contextualizes her work, but not whether or not she should publish it. In “Overcoming Obstacles in Quantitative Feminist Research” (this issue, pp. 419–426), Apodaca reflects on the politics of data—what is collected, what is not collected, and the challenges of interpreting data with variables that are measured differently across countries and over time, or when the conceptualizations of those collecting the data embed meaning and interpretation in the data. “For example, for a government agency collecting labor data, women's work may not be considered work but simply ‘chores’” (pp. 419–426). Or consider Caprioli's long struggle to find a measure for gender equality that gives her confidence that her variable is a good operationalization of the concept of gender equality (this issue, pp. 426–431, and Caprioli Reference Caprioli2000). Such frank discussions about wrestling with politics while we study it makes these contributions particularly engaging reading and, hopefully, useful in assessing existing feminist empirical work and designing future research.

All three struggle with operationalizing concepts as variables for which they had or could get data. In her early work, Caprioli uses an operationalization for which there was weak theoretical and empirical evidence (Caprioli Reference Caprioli2000). Now she has data that offers empirical support for the earlier choice (Caprioli et al. Reference Caprioli, Hudson, McDermott, Ballif-Spanvill, Emmett and Matthew Stearmer2008).

Parisi, Apodaca, and Caprioli have widened the terrain of feminist methodological discussion to include quantitative methods. They open at least three doors that invite further debate within the field: The structure of the discipline, the construction of data sets, and the importance of feminist analysis of the underlying data themselves.

First, as True and I have argued in this journal, the field of IR is socially constructed in a way that reinforces norms of professional achievements by predominantly U.S. IR scholars exploring questions using positivist methods (Ackerly and True 2008). Caprioli notes the influence of these norms on her graduate training. Parisi struggles against these norms when she thinks about publication choices and her tenure prospects. Apodaca consciously works within them with the expressed purpose “to empower women and create social change.”

Second, feminist quantitative methodologists are constrained by existing data sets or become engaged in creating them. The research agenda of reconceiving the conceptual building blocks of IR—state and security, for example—cannot be pursued using quantitative methods alone if the data available are collected in ways that reify those building blocks. Among others, gender disaggregated data is necessary. As Apodaca notes, creation of such data sets can reveal discrimination and itself be a political act. Apodaca, Parisi, and Caprioli have made the development of such data sets a significant part of their research, as have others (True and Mintrom Reference True and Mintrom2001).

Third, like qualitative data, quantitative data needs to be analyzed. As indicated before, feminist analysis is used to decide what measures to use for “women's human rights” or “gender equality,” for example. Feminist analysis can tell us the merits and shortcomings of possible variables for measuring the concept in question. Feminist analysis is used to interpret the data. For example, feminist theory tells us to study intersectionality, the possible differences among women by race, age, and other variables. As Apodaca notes, in quantitative analysis, interaction terms can enable us to study some aspects of intersectionality. Additionally, feminist analysis can tell us what other variables are necessary. For example, in early work, Caprioli uses fertility as a measure of women's social equality. Since high fertility is also connected with poverty and infant mortality, we need to control for these variables in quantitative models in order to capture the “gender” effect of fertility. Feminist analysis can reveal systematic biases in data, as in Parisi's analysis of the use of the male norm in much human rights data.

Certainly, there are tensions within feminism such that there is not a shared view on any of these matters, and feminists need to work among themselves while engaging with the mainstream to improve the quality and exposition of our research. However, as we work, we do not want feminist inquiry (or any inquirer's research) to become collateral damage. Critical engagements among feminists improve the quality of empirical scholarship. Engagements by non-feminist-identified researchers with the questions and approaches of feminist empiricists can strengthen the field as well. Certainly, the work of scholars preferring to use quantitative methods for pursuing feminist questions belongs in both feminist and mainstream IR publications.

I began by recognizing the importance of empirical work for feminist theory; let me close with reflections on the value of feminist theory for empirical work. First, feminist theory facilitates the conceptualization and reconceptualization of the ideas we wish to study. Second, feminist theory bridges philosophical methods and social science methods, adding to the rigor of our scholarship. These are the two that feature most prominently in the essays by Parisi, Apodaca, and Caprioli.

A third is that which Tickner has in mind in her engagement with Keohane: Feminist normative theory tells us why a particular research question is important. Different feminist theories will suggest that different empirical questions are more important, for example, the conceptual building blocks of the state (Peterson Reference Peterson1992; Tickner Reference Tickner2001), political economy (Peterson Reference Peterson2003), postcolonialism and the deconstruction of empire (Ling Reference Ling2002; McClintock Reference McClintock1995), and nationalism (Dhruvarajan and Vickers Reference Dhruvarajan and Vickers2002; Kaplan, Alarcón, and Moallem Reference Kaplan, Alarcón and Moallem1999). There are many differences among feminists beyond our different questions and methodologies—normative and empirical, qualitative and quantitative. Often, feminist theory debates the differences, the extent, and the kind, as well as whether we are adequately attentive to all of them. When we study the frank reflections of scholars attempting to be theoretically and methodologically rigorous—even to redefine what we mean by rigor and whether rigor is the right measure of good scholarship—we can learn from our learning. When empiricists share in publication the challenges that we normally share only with those whose feedback we have solicited, we enrich the landscape of methodological debate that can inform both empirical and theoretical work.