Feminism Counts

Quantitative Methods and Researching Gender

Christina Hughes, Rachel Cohen, Christina Hughes, Rachel Lara Cohen

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781317986218
• Edition: 1
• Topic: History
• Subtopic: Social History
• Index: History

This is an important and timely text that provides a unique overview of contemporary quantitative approaches to gender research. The contributors are internationally recognised researchers from the UK, USA and Sweden who occupy a range of disciplinary locations, including historical demography, sociology and policy studies. Their research includes explorations of heterosexual and same sex violence, media responses to feminist research, data sources for the study of equalities, approaches for analysing global and local demographic change and intersectional concerns in respect of work and employment.

Through detailed, sophisticated and thoughtful considerations of the place of quantification within gender studies, and the place of feminist approaches to quantification, each contributor overturns the stereotype that quantitative research is antithetical to feminism by demonstrating its importance for challenging continuing global inequalities associated with gendered outcomes. An introductory chapter illustrates the significance of geography and discipline in the take-up of methodological preferences.

Feminism Counts: Quantitative Methods and Researching Gender makes an important contribution to the ways in which feminists respond to contemporary methodological and interdisciplinary challenges, and is essential reading for all research students in gender studies.

This book was originally published as a special issue of the International Journal of Social Research Methodology.

Feminists really do count: the complexity of feminist methodologies

We are delighted to present this special issue on feminism and quantitative methods. We believe that such an issue is exceptionally timely. Within the field of feminism and women’s studies debates about, and resistance to, quantification persist. These, however increasingly intersect with debates within the wider research community about the development of appropriate methodologies that take account of new technological and philosophical concerns and are fit-for-purpose for researching contemporary social, philosophical, cultural and global issues. Two areas serve as exemplars in this respect as they combine wider social science and specifically feminist methodological concerns. The first is the increasing focus amongst social scientists with capturing and analysing the complexity of social life. Within feminism, this is seen in debates about intersectionality and the need to recognise the implications of multiple social positions/divisions and associated power structures. As Denis comments in respect of intersectional analysis methodological development is struggling to catch up with theory:

The challenge of integrating multiple, concurrent, yet often contradictory social locations into analyses of power relations has been issued. Theorising to accomplish this end is evolving, and we are struggling to develop effective methodological tools in order to marry theorising with necessary complex analyses of empirical data. (2008, p. 688)

The second example of overlap in the methodological ambition of feminist and general social science is related to new techniques and data sources coming on line. For instance, in the UK the ESRC National Data Strategy has set out priorities for the development of research data resources both within and across social science boundaries that will facilitate historical, longitudinal, interdisciplinary and mixed methodological research. These developments may thereby facilitate a longstanding feminist aim, not simply epistemological and methodological interdisciplinarity but transdisciplinarity.

Debates within second wave feminism have, however, left their legacy. Here, sustained critiques were developed of how, for example, quantitative methodological tools objectified subjecthood; how objectivity itself was a smokescreen for male interest, male perspectives and male privilege; and how ‘woman’ (both literally in terms of research respondents and in terms of epistemological foci) was missing from much research (see, e.g., Mies, 1983; Oakley, 1981; Stanley & Wise, 1993). Whilst the broader ‘qualitative versus quantitative’ paradigm debates predate these critiques, feminism added a dualistic critique that associated the binary of qualitative and quantitative with a dualistic structuring of female/male; soft/hard; intuitive/rational; art/science and so forth (Oakley, 1998). In consequence, feminists called for a feminist methodology that, as the slogan at the time reflected, would not be ‘on women’ but would be ‘for women’ and later, as participatory approaches developed would be ‘with women’. Qualitative methodologies came to the fore as most suited to this approach because they were viewed as the most appropriate way of enabling researchers to listen to and give voice to women. Over time this became orthodoxy: feminists use qualitative methods. Thus Oakley (1998, p. 716) notes that ‘although there are some signs of a new recognition within feminist social science of the usefulness of non-qualitative methods, both feminist methodology and feminist epistemology remain strongly founded on qualitative methods’.

When Oakley (1998) tried to rehabilitate quantitative methods, she aroused considerable disquiet within feminism, in part because her concern with ‘usefulness’ had pragmatic overtones. There is, however, a pragmatic politics to be considered in terms of the efficacy of quantitative approaches for feminist aims for social justice. Such pragmatics recognises that quantification remains the ‘gold standard’ for much social science and policy-oriented research, despite ongoing critiques from feminist researchers at an international level about public funding authorities’ ‘preference’ for quantitative approaches rooted in realist and positivist frameworks (see Davies & Gannon, 2006), and despite the massive growth, and seeming greater acceptance of qualitative approaches. Whilst the division between qualitative and quantitative research within and without feminism persists, it reinforces a host of gendered, and counter-productive notions around hard/soft; emotional/rational; worthy/worthless dualities. Given the significance of quantified and quantifiable data to policy deliberations associated with social justice, it is crucial that we escape these simplistic dichotomies and re-open quantitative methodologies to critical feminist epistemological and feminist empirics.

It is within this terrain that this collection seeks to make a contribution. Our aims are to:

• Contribute to and invigorate debates concerned with feminist approaches to quantitative methods;

• Contribute to widening the understanding and value of feminist approaches to quantitative methods;

• Inform debate about the contemporary concerns of those using quantitative methods for gender research.

In setting out the parameters of this volume, we begin by asserting that there is strong support for the belief that qualitative approaches predominate within what we define as the broader field of feminism. However, the contents of this special issue and our own research – which draws on an analysis of articles published in gender, women’s studies and feminist journals – challenge the simplistic, and consequential, presumption that to do feminist empirical research one has to use qualitative methods. Our own analysis (see Cohen, Hughes, & Lampard, 2009 for a fuller account) indicates that we need to add two further factors into any argument about methodological dominance or preference. These are geography and discipline. In terms of geography, there appears to be a pattern of US exceptionalism in terms of feminist or women’s studies methodological approaches in English language journal publishing. In our analysis we used the national base of the first-listed author as a simplified proxy for geographical location. Of the articles, we sampled with a first-author based in the USA, 60% were quantitative and 41% qualitative, something that was reversed for articles with a UK-based first-author (with figures of 13% and 73%, respectively). Articles with first-authors from other countries were, like those with a UK-based first-author, much more likely to be qualitative. Thus, as suggested by the differences found in British and American sociology (Dunn & Waller, 2000; Payne, Williams, & Chamberlain, 2005; Platt, 2007), geographically specific norms impact upon methodological choice with researchers from the USA much more likely to draw on quantitative approaches than other countries. Our analysis also demonstrated the importance of discipline and, associated with this, the journal within which articles appear. Thus, the journals Feminist Economics, Women’s Health Issues and Psychology of Women Quarterly evidence a strong quantitative dominance in line with the disciplines of economics, medicine and psychology. In comparison, there was a lack of any quantitatively based articles in our sample of interdisciplinary Gender or Feminist Studies journals: Feminist Studies, Social Politics and Journal of Gender Studies. Therefore, where feminist or gender analysis occurs within distinct disciplinary contexts, methodological choice is influenced by that context. Yet articles in more general feminist or women’s studies journals generally conform to the widely recognised feminist ‘qualitative bias’. This suggests a possible tension – between feminist scholars working within disciplinary boundaries and the larger feminist project of transdisciplinarity. Additionally, it may indicate a certain practical pragmatism whereby feminist scholars in primarily ‘quantitative’ fields have negotiated their own compromises between disciplinary norms and practices and the feminist critical project. There has, however, been little public discussion or methodological consideration of how this has occurred, which is why we believe that the discussion of feminism and quantitative methods is long overdue.

The issue of disciplinary location is at the forefront of the concerns of two authors in this volume (Jill Williams and Lotta Vikstrom). Both authors in different ways reprise earlier feminist critiques of the reductionist and marginalising epistemological tendencies of a discipline. As such, these pieces are important reminders that we should not rush toward a new methodological future without recognising that there remain some longstanding and seemingly intractable methodological concerns about the priority given to quantitative findings and that second-wave feminist critiques retain their potency. Jill Williams clearly articulates how the quantitative disciplinary approaches of demography resist feminist knowledge. As Jill notes, this is despite the discipline’s concerns with issues of empowerment and women’s lives globally, and the appeal of these foci to feminists. She highlights how a quantitative approach that regards analysis as primarily concerned with the manipulation of key variables results in a focus on power in gendered relationships dropping out of view. Within this framework, qualitative approaches are acceptable only insofar as they provide technical support to quantitative approaches, a route to improving the validity of measures. Jill critiques the practice of demographers to simply treating ‘Gender’ as an independent variable, invariant across time or place, given extensive feminist scholarship that demonstrates that the category ‘woman’ (and similarly ‘man’) is socially constructed and differently salient in different contexts. Her suggestion is that demographers treat gender as a dependent variable, examining the production of gendered inequalities. She convincingly argues that this would fit with demographers’ interest in content validity, as the suggestion is effectively a call to unpack the meaning of gender. In addition, Jill suggests that feminist reflexivity over issues such as the politics of location would enable demography to counter its tendency to Western bias both in how research questions are framed and the lenses brought to data analysis.

Lotta Vikstrom also focuses on demographic research, although in this case historical demography is the field of enquiry. She notes that few feminists work in historical demography, suggesting that this may be because of its quantitative orientation. A contrast is drawn with social and cultural history, a field in which feminism has made a significant impact. Lotta argues that consequently historical demography lacks a feminist perspective and that there is a strong divide between quantitative and qualitative scholars. She outlines an interesting case study that attempts to bridge this gap. The study is an exploration of the occupations of women living in Sundsvall, Sweden during the period 1860–1890. Lotta demonstrates the importance of digitisation in enabling researchers to access a variety of sources: trade directories and registers; census data; newspaper reports. This highlights the potential of new digital technologies, suggesting that these may not simply facilitate but may begin to require researchers to work across long policed boundaries, combining quantitative and qualitative approaches. By bringing together data from multiple sources on the same women Lotta exposes the problems and issues that arise when the data are dissonant, contradictory or indeed confirmatory. In this sense, Lotta demonstrates a concern that has been central to feminist scholarship – to retrieve the actual lives of individual women from the historical records. She argues that ‘triangulation’ enables the researcher to investigate data validity and explore the problems associated with paradigmatic boundaries.

The issue of paradigmatic boundaries and mixed methodologies is a theme of many authors in this volume. This is perhaps most strongly expressed by Jacqueline Scott who explicitly states that there is no one best method, rather different methods are appropriate for addressing different research questions. She argues that, importantly, quantitative analysis provides evidence of the modification or reproduction of gendered inequalities over time and space. Jacqueline draws on studies being undertaken as part of the ESRC Research Priority Network on Gender Inequalities in Production and Reproduction (GeNet), which explore the inter-relations of paid and unpaid work. She demonstrates the contribution made by feminist critiques of the false divide between public and private and the recognition of intersectionalities to methodological developments including the kind of triangulation indicated by Lotta Vikstrom. As Jacqueline notes, however, these forms of analysis are still in their infancy in part because of data limitations. For example quantitative approaches to intersectional analysis require very large samples. Similarly, exploration of gendered power in time allocation requires large time-use studies, which allow family and individual time to be compared. Given the somewhat patchy availability of data the significance of a methodologically and disciplinary diverse network such as GeNet is all the greater. Like Jill Williams, Jacqueline’s emphasis is on gender as a historically and socially specific category. She demonstrates how interchanges of quantitative and qualitative research amongst network participants can move us away from outdated, static and universalising accounts of gender inequalities to more nuanced and contextualised findings that recognise the complexities of social inequalities. Overall she makes a convincing argument that good quantitative analysis, is not ‘simply gathering numbers’, but fundamentally rooted in developing a theoretical understanding of gender.

Jacqueline Scott’s contribution is also significant for its remarks on the issues of capacity building around skill and knowledge, particularly in terms of quantitative approaches, that will be required if feminists are going to be able to rise to the methodological challenges we currently face. Sylvia Walby and Jo Armstrong’s piece is especially pertinent in this regard. They present some of the work they have undertaken for the UK Equality and Human Rights Commission (EHRC) on developing measures of equalities. Sylvia and Jo note the contentious issues associated with measuring such a complex concept as equality and yet recognise the importance of confronting these dilemmas if we are going to improve comparative, (trans)national and intersectional analysis, making it accessible to wider audiences, and importantly, speak to policy-makers. They outline their framework, which includes issues of definition and data availability and the technical aspects of measuring equality. They argue that the criteria for judging fitness for purpose are: relevance, accuracy, timeliness, accessibility and clarity, comparability and coherence. They conclude with what they refer to as ‘Headline’ indicators which they suggest, until more data and greater harmonisation of data occurs, provide for a robust assessment of the degree of equality across the economic, political and social spheres.

In detailing their approach to developing these ‘Headline’ indicators, Sylvia Walby and Jo Armstrong note the struggles that ensue over definitional criteria. This is because measures can have material consequences. What might appear to be small details can make a considerable difference to revealing or obscuring levels of inequality and thereby prompting or delaying policy change. Marianne Hester, Catherine Donovan and Eldin Fahmy demonstrate these concerns in terms of the development of survey instruments in the field of domestic violence and same-sex relationships. They provide a further demonstration of the contribution made by debates within feminist epistemology to the development of more sensitive methods for understanding social concerns. In this regard they note how similar survey instruments, principally the Conflict Tactics Scale (CTS), have had been used to study domestic violence in both heterosexual and same-sex relationships. Thus, some of the critiques of surveys focused on heterosexual relationships also apply to those concerned with same-sex relationships. Critically, the CTS measures prevalence of violence but pays little attention to impact. Additionally, since qualitative studies have shown that the impact of domestic violence is gendered the authors suggest that it is crucial that a survey instrument is able to examine how this may differ between same-sex and heterosexual couples. Marianne, Catherine and Eldin take a feminist epistemological approach to gender and power, problematising the relationship between experience and discourse. This allows them to redevelop the CTS and to disaggregate the experience of abuse, its impact and the perception of that experience (whether respondents themselves believed that they had been experienced domestic abuse). In doing this Marianne, Catherine and Eldin provide a robust argument that an assessment of severity of impact requires these multiple measures. Moreover, like Jill Williams, they illustrate the utility of a feminist epistemological approach for the production of better quantitative analysis.

As Marianne, Catherine and Eldin note geography as well as discipline affects methodological direction. In this respect, they confirm our own analysis that North American approaches are more heavily dominated by quantitative than qualitative approaches. In the USA, for example, there has been a burgeoning of large-scale quantitative studies of domestic violence, which use comparative scales and inventories. In the UK and Europe, the focus has more strongly been toward the analysis of personal experience via phenomenological and critical research. Interestingly, however, Marianne, Catherine and Eldin also point to the ways in which gender influences methodological approaches more absolutely with research on gay men more likely to be quantitative whilst research on lesbian relationships has been more qualitative. Diane Croker, discussing Canadian surveys on woman abuse, also confirms the predominance of quantitative approaches to issues of gender and violence in North America. However, Diane’s contribution is less concerned with definitional issues of abuse or methodological precision. Indeed, she suggests that we have collected sufficient data about the levels of abuse women are subjected to. Rather, she analyses the constitutive role of statistics and the consequences of their reception for policy development. Diane draws on Foucauldian perspectives of governmentality to analyse how woman abuse is publicly debated in neo-liberal societies. She compares the reception of two surveys: Violence Against Women Survey and The Dating Violence Survey. Her analysis describes differences in the responses to these surveys from feminist, anti-feminist and media sources. Significantly, however, Diane argues that whilst definitional debates and struggles continue, surveys that measure individual experiences of abuse both construct such abuse as ‘an injurious act’ and focus policy toward criminal justice solutions concerned with punishing perpetrators rather than toward policies that would better support collective and individual women in the multiplicity of ways, in which they seek to resist violence or, indeed,...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Contents
• Notes on Contributors
• 1. Feminists really do count: the complexity of feminist methodologies
• 2. Doing feminist-demography
• 3. Identifying dissonant and complementary data on women through the triangulation of historical sources
• 4. Quantitative methods and gender inequalities
• 5. Measuring equalities: data and indicators in Britain
• 6. Feminist epistemology and the politics of method: surveying same sex domestic violence
• 7. Counting woman abuse: a cautionary tale of two surveys
• Index