Feminist Methods SHULAMIT REINHARZ with the assistance of LYNN DAVIDMAN New York Oxford OXFORD UNIVERSITY PRESS 1992 196 Feminist Methods in Social Research in any formal cost-benefit analysis of the tradeoffs between employment and family life.90 The voices of feminist scholars engaged in different forms of action research th include the individual who honestly assesses what she has learned about herse By including this perspective, I find a strong connection between the activism feminist action research, and the self-reflexive nature of much feminist resear that does not label itself as activist. For this reason, although a chapter on acti. research is useful in stressing change-oriented forms, it would be misleading if suggested that other forms of feminist research are static. To the extent that fei inism is change-oriented by definition, all feminist research has action comp nents. Feminist Multiple Methods Research An emerging postulate for feminist research—is using a variety of methods in order to generate multifaceted information.2 Just as feminist research often draws on multiple disciplines, so too it often draws on multiple, rather than a single, method in a particular project. By contrast, many studies in the social sciences do not use multiple methods but rely on one data type or method of analysis. The use of multiple methods in a single study has earned its own name—triangulation.3 Although like mainstream research, feminist research is frequently unimethodological, there are many instances where it is triangulated. There may be a greater proportion of triangulated feminist research than mainstream research because of the special relation triangulation has with feminist concerns. Feminists choose multiple methods for technical reasons, similar to mainstream researchers, and for particular feminist concerns that reflect intellectual, emotional, and political commitments. Feminist descriptions of multimethod research express the commitment to thoroughness, the desire to be open-ended, and to take risks. Multiple methods enable feminist researchers to link past and present, "data gathering" and action, and individual behavior with social frameworks. In addition, feminist researchers use multiple methods because of changes that occur to them and others in a project of long duration. Feminists describe such long projects as "journeys." Sometimes multiple methods reflect the desire to be responsive to the people studied. By combining methods, feminist researchers are particularly able to illuminate previously unexamined or misunderstood experiences. Multiple methods increase the likelihood of obtaining scientific credibility and research utility. Historical Roots Agnes Riedmann argues that Margaret Loyd Jarman Hagood (1907-1963), a pioneer in statistics and demography, should be recognized as an early advocate of combining qualitative and quantitative methods in sociological research.4 After earning an M.A. in mathematics, Margaret Hagood earned a Ph.D. in sociology. 197 198 Feminist Methods in Social Research Her 1936 doctoral dissertation consisted of a statistical analysis of the fertilit patterns of white women in the rural southeast of the United States, then th region of highest population growth in the country. In her first position as a research associate in the Institute for Research i Social Science at the University of North Carolina, Chapel Hill, Margaret Hagoo reexamined the same population she had studied in her dissertation. This time sh also traveled throughout the south and interviewed farm tenant wives at length Her resulting book, Mothers of the South, was a well-received qualitative studs She wrote: It is already well established that rural women in the South have a high level of fertility and a low level of living, but little is known of what these quantitative measures mean in the lives of the mothers and their children. Therefore, the effort was made to be both more realistic and comprehensive in a sort of attempt to get away from restricted statistical treatment and stereotyped "case studies," and yet to combine as far as possible some features of the statistical, case and survey methods. Preliminary work included first a statistical analysis of the regional and subregional variations in fertility, which served to locate and define certain groups of high fertility levels, and to examine the implications of fertility differentials with particular reference to the Southeast. . . . The actual first-hand gathering of data consisted of repeated visits made by the writer to the tenant farm mothers during a sixteen months' period of field work.3 At the same time, she maintained her interest in statistics and demography and published the very popular text Statistics for Sociologists in 1941.6 Margaret Hagood expertly combined quantitative and qualitative methods in her studies of the life experiences of women and families in the rural south. She personally interviewed members of 254 homes at least once, choosing the households on the basis of indices of mobility, education, age, age at marriage, number of years exposed to pregnancy, fertility, and occupation. The farmers were able to speak freely with her because they saw her as only slightly removed from their life-style, having been raised in the same area. One of her interesting findings is that seven of eight of the women she interviewed preferred to work in the field rather than in their homes. They were prouder of their accomplishments outside the home than inside. At the same time she found that the women and their families had inadequate diet, housing, medical treatment, education, social life, and recreation. The major culprit was "the burden of involuntary and frequent childbearing," which in turn was aggravated by lack of regional planning, inadequate state contraception programs, overreliance on a single crop system, and disregard for this region by the federal government. Margaret Hagood subsequently made use of another method— photography—which I will discuss in the next chapter on "feminist originals." For more than 50 years, sociologist Mirra Komarovsky has identified herself as a feminist and has conducted multiple method research.7 In her latest major study, funded by the National Institute of Education and various private foundations, she offers a model of multiple methods research that integrates methods within sociology, and combines these with psychological instruments. Her study Feminist Multiple Methods Research 199 documents how women change when educated in an elite women's college in the United States: The study began in the fall of 1979 when 630 freshmen entered college. A random sample, stratified by race and religion, was selected out of the total. It consisted of 241 students who each received a set of questionnaires, scales, and tests. By my and my assistant's deadline, two weeks later, the rate of return was 96 percent of the original number approached, giving us a sample of 232 students, including both residents and commuters. Following the distribution of self-administered questionnaires and scales, Mirra Komarovsky used interviews and diaries: 201 students were interviewed for about an hour, to clarify and supplement their completed forms. ... In the spring semester, nine freshmen kept diaries, following an outline I prepared, and were interviewed every two weeks. (They received a modest honorarium for this work.) The purpose of the diaries was to generate hypotheses regarding the impact of college on the study's dependent variables for the interviews in the fall of 1980. In keeping with this longitudinal design, in the fall of their sophomore year "students received replications of all the research instruments measuring dependent variables." In addition, "a subsample of 70 students agreed to two two-hour interviews." The students were not studied during their junior year, but as seniors, materials replicating those sent them in their sophomore year were collected. Sixty-five seniors consented to two-hour interviews. As Mirra Komarovsky summarized, Our methods . . . combined quantitative measures (questionnaires and scales) and qualitative data (interviews and diaries). Although the interview was the major research tool, the interplay of both methods was a distinct feature of the study. Mirra Komarovsky does not draw on feminism to explain her use of multiple methods. In fact, in other publications she rejects the idea that there is a particular feminist methodology. At the same time she advocates strongly that data be interpreted in feminist ways. Mirra Komarovsky regards interviews as providing internal validity for the statistical patterns uncovered by surveys. At the same time interviews convey additional information about experience unobtainable in surveys. She explains her inclusion of excerpts from the interviews throughout the book as a "kind of fusion of scientific and iiterary' functions, to convey the immediacy of an experience even as it seeks to communicate some of the major theoretical concerns of this study."8 Finally, she expresses the hope that this combination of data will make her book even more useful to students, professors, college personnel, parents, and policy makers than it would be if based on one type of data alone. Thus, in addition to providing a model of personal continuity in multimethod research, Mirra Komarovsky illustrates two rationales for doing this type of research: to enhance its scientific status and increase its potential utility to readers. 200 Feminist Methods in Social Research Rationales for Contemporary Feminist Multimethod Research Triangulation of Quantitative and Qualitative Methods An article concerning sexual harassment, by psychologist Bernice Lott and her colleagues Mary Ellen Reilly and Dale Howard, is an example of a straightforward triangulation of quantitative and qualitative materials. In the following excerpt they explain the items in their questionnaire, the nature of their interviews, and the fact that they provided counseling services on request: The questions in our nine-page survey . . . were applicable to the experiences of both women and men and covered the following areas: demographic information (age, university status, etc.); knowledge of other persons on the campus who had been sexually assaulted; personal experience of assault on campus; personal experience of sexual assault anywhere; actual and potential experience as sexual assaulter; knowledge of sexual intimidation experienced by others on campus; personal experience of intimidation on campus; experience of having been offered sexual contact in exchange for job- or school-related benefits; opinions about the frequency of sexual insults on and off campus; personal experience of sexual insult; and finally, attitudes toward and acceptance of specific sexually harassing behaviors .... A separate sheet of paper was included with each survey on which a respondent could volunteer to be interviewed. We also supplied the names and phone numbers of campus counselors available to anyone who wished "to talk confidentially with someone about any of the issues raised" in the survey. Persons who indicated a willingness to be interviewed were contacted individually by one of five interviewers familiar with the objectives of the investigation and the contents of the questionnaire. Each interview was conducted by a same-sex interviewer. . . . Each interview was open-ended and began with the simple question "What would you like to tell me about the issues dealt with in the survey?"9 In sum, their study consisted of distributing surveys following by open-ended interviews with a subgroup of volunteers, each method fleshing out the other. Sociologist Pauline Bart and her colleagues Linda Freeman and Peter Kimball studied attitudes toward pornography in a multimethod manner. Their first step was to obtain preliminary data in a natural setting. They did this by handing out stamped, addressed, pre-folded questionnaires to patrons exiting a theater showing the film (Not a Love Story). This questionnaire asked only for the gender of the respondent. After telling them the importance of the issue, they were asked to cooperate by writing their reactions to the film including any changes it had made in their attitudes about pornography, what they liked and didn't like about the film, what they learned, if anything, and anything else they wanted to tell us. Another data collection procedure in the field was to interview people who had seen the film. Based on both sets of information, they produced a questionnaire including forced-choice and open-ended items answered by 668 people who had seen the film in the recent past. Pauline Bart and her co-researchers explained the questionnaire and the way the responses were analyzed: Feminist Multiple Methods Research 201 Each of the 42 items on the questionnaire was taken directly from the answers. The statements were about equally divided between pro-and anti-pornography attitudes. The subjects were asked to indicate degree of agreement with each statement, ranging from "strongly agree" to "strongly disagree." There was no zero point. Standard demographic questions were included, as well as a question asking respondents if they were feminist or pro-feminist. . . . We also included a blank space on the questionnaire, which they were told they could use to say anything they wanted. . . . The answers to these 42 pre-coded questions were cross-tabulated by gender and by type of feminism or pro-feminism with gender controlled. . . ; The written comments were analysed in two ways. First, Freeman examined all the comments and constructed a typology. In order to have a more rigorous qualitative analysis, a second classification was then constructed, independently of the Freeman classification, which could account for each response.10 Thus this project conducted in-the-field interviews in order to develop a questionnaire, and produced an integrated analysis of quantitative and qualitative data concerning people's reactions to a specific film that deals with pornography. Unfortunately, in mainstream research projects open-ended interview data are sometimes collected to supplement surveys, but then are not analyzed.11 Commitment to Thoroughness Feminist researchers combine many methods so as to cast their net as widely as possible in the search for understanding critical issues in women's lives. The multimethod approach increases the likelihood that these researchers will understand what they are studying, and that they will be able to persuade others of the veracity of their findings. Multiple methods work to enhance understanding both by adding layers of information and by using one type of data to validate or refine another. Psychologist Phyllis Chesler's study about women and madness is an example of a feminist research project that turned to multiple methods because of her commitment to examine thoroughly a phenomenon that was poorly understood. In the introduction to her book, she explains that her material begins with the presentation of the lives and "psychiatric" histories of four women, based on autobiographical, biographical, and "case history" material. These, and modern women in general, are viewed in terms of what growing up female in the family means. . . . Chapter One describes how female reproductive biology, patriarchal culture, and the modern parent-daughter relationship have so combined as to insure such characteristically female behaviors—and ideals—as self-sacrifice, masochism, reproductive narcissism, compassionate "maternality," dependency, sexual timidity and unhappiness, father-worship—and the overwhelming dislike and devaluation of women. After this discussion of case studies, she analyzes "both the mental asylum and private therapy as recapitulations or mirrors of the female experience in the family" and then reviews traditional and contemporary clinical theories and practices. 202 Feminist Methods in Social Research Although Phyllis Chesler is a clinical psychologist, she also deals with such • ciological materials as "our nation's mental illness' statistics from 1950 to 1959 In the largest section of the book, she presents her clinical materials. These elude the "patient 'careers' of sixty women whom [she] interviewed about th experiences in psychiatric hospitals and private or clinic outpatient therapy—^ periences which took place from 1945 to 1971 in America and England." ln \ concluding comment, she explains that she speaks in "many voices . . . & psychological researcher, theoretician, and clinician—and as a literary and phi sophical person, a lover of poetry and myths."l2 In my view, Phyllis Chester's last comment indicates that the use of multiple methods reflects the multifaceted identity of many feminist researchers. We are multifaceted because we are working during a feminist renaissance that transcends disciplinary boundaries and challenges many of our capacities at once. Our mul-tifacetedness makes single-method research seem flat and inadequate to explore and express the complexities of women's lives. As did Phyllis Chesler, Carolyn Sachs combined historical and contemporary materials in her study of women farmers. In addition to utilizing archival documents from the seventeenth century to the present, she analyzed changing conditions of production and the modification of ideologies concerning the sexual division of labor. To supplement these materials, she conducted in-depth interviews with 21 female farmers so that the women could speak for themselves and suggesi ways of eliminating the subordination under which they suffered. Her multiple method research seemed to be motivated by a desire to connect women of the past and present. Similarly sociologist Ruth Wallace combined analysis of documents from Catholic institutions with interviews of Catholic women recently appointed as church administrators in her study of women in the Church hierarchy,13 and Leila Rupp (historian) and Verta Taylor (sociologist) combined archival analysis and interviews to study the U.S. women's movement in the recent past.14 Judith Herman and Lisa Hirschman's illuminating study of incest also used a multiplicity of methods. As feminist clinicians, they were alarmed by the number of women they treated who were incest victims. The lack of available literature on the topic of father-daughter incest inspired them to write their book. They felt driven to be as thorough as possible because of the horrific quality of the phenomenon they investigated and the responsibility they felt to incest survivors. They also believed that they would best be able to fill the vacuum of knowledge on incest by drawing on all the methods available to them. The use of multiple sources of data, I contend, reflects the passion they brought to the study. At the outset they explain that they divided their book into three parts, each containing different kinds of materials: survey data, clinical material, anthropological literature, popular literature, and pornography. For our scholarly sources we naturally relied on libraries; but for much of the popular literature we relied on the help of friends, acquaintances, and strangers who clipped newspaper or magazine articles and called our attention to publications that we would not ordinarily have seen. Following this overview of relevant materials, they include "a clinical study of [their] own, based upon interviews with patients in therapy or their therapists. Feminist Multiple Methods Research 203 ;orty incest victims and twenty women whose father had been seductive but not vertly incestuous participated in our research." Feminist theory (in this case, linking concepts of male supremacy and female ppression) enabled them to make sense of these multiple sources of information. !ut they were not interested in confining their research to analysis and theory uilding- They were also interested in action. Thus, one part of their book deals núl the social responses to discovered incest, including crisis intervention, family treatment, and prosecution. It also deals with the possibilities of healing and prevention. Our discussion is based upon interviews with professionals in the mental health system, child protective services, and law enforcement, and upon site visits to programs that have developed innovative approaches to the problems.15 [n other words, they rounded out their study of the literature and actual clinical cases with an investigation of the social context of incest and a suggestion for preventive action. Sociologist Lenore Weitzman's study of "the social, economic, and legal consequences of California's no-fault divorce law" required a set of methods for each of these three consequences. As she explained in her appendix on research methods: The research involved four types of data: systematic random samples of court records; interviews with members of the matrimonial bar; interviews with superior court judges and commissioners who hear family law cases; and, perhaps most important, interviews with recently divorced men and women .... In all of the interviews that we conducted, each respondent (i.e., each judge, lawyer, divorced man and woman) was presented with a series of hypothetical divorce cases (in addition to about one hundred pages of questions). Four of these cases are discussed throughout this book: one involves a young couple with preschool children, another concerns an older corporate executive and his wife after a twenty-seven-year marriage, the third focuses on a middle-aged nurse who supported her doctor husband through medical school, and the fourth deals with noncompliance with a child support award. Seeking to avoid U.S. ethnocentrism, she reports that she had the opportunity to conduct a similar set of interviews in England and to explore the responses of English judges and attorneys to the same four hypothetical cases. Their reactions serve to highlight the unique features of the California legal perspective.16 The records of divorce cases her research team examined were drawn from samples before and after the implementation of the no-fault law and in different counties. All told, the multiplicity of her methods and the skillful manner in which her data were analyzed enabled her to demonstrate the extremely harmful and inequitable effects of the no-faull divorce law on women. Integration of the Personal and the Social Kathleen Barry's study of prostitution and the traffic in women, phenomena she labeled "female sexual slavery," required multiple methods because the activities 204 Feminist Methods in Social Research are clandestine and multifaceted.17 Her very disturbing book begins with a harrowing ethnographic description of a particular form of prostitution in Paris that she was able to observe. Other methods she used include gathering archival material, analyzing documents, sitting in on court trials, and using investigative reporting techniques to identify individuals involved in the traffic. Even after locating women who had escaped from or left prostitution, she found it difficult to interview them because they were convinced she would not believe them. To increase their trust in her, Kathleen Barry conducted lengthy open-ended interviews that enabled the women to speak freely and at length. By encouraging them to describe their experiences accurately, she increased their self-confidence while avoiding exaggerations in their stories. Although she relied on interviewing for its particular assets, Kathleen Barry was careful not to analyze prostitution as the "personal problem" of the particular woman. Rather, she was interested in the conditions that had enslaved these particular women as prostitutes, and that could have had that influence on any woman. In other words, her multiple methods helped her link the case of the individual woman with a broader complex of social and economic issues. Multiple methods are used by many feminist researchers because of our recognition that the conditions of our lives are always simultaneously the product of personal and structural factors. The Long Duration of Research Projects Another rationale feminist researchers offer for using multiple methods is the long time period needed to complete an in-depth study. In that duration, we may learn new ways of doing research or may discover that the circumstances of the people we are studying have changed. An example can be found in Cynthia Fuchs Epstein's research on women in the law. She writes: This analysis of the changing role of women in the legal profession is based on fifteen years of research using a number of different "methods." It is hard to isolate method from general experience when one has spent so much attention on a research subject for a long period of time. ... I changed methods in the course of those years, partly because it seemed appropriate to my goals and to the changing situation of lawyers. I was also exposed to different styles of work in sociology, work that did not replace earlier perspectives that had formed my sociological vision but that, I felt, added a richness and depth.18 Breaking down the barrier between her life and her work, Cynthia Epstein describes research as a fluid, flexible process that takes on different methods in a responsive way. Betsy Ettore's study of "the London lesbian ghetto" demonstrates how different methods became available to her as the people she was studying began to trust her. Spending a lot of time doing a project communicates seriousness and commitment to the people one is studying: Feminist Multiple Methods Research . 205 During the beginning of my research I was able to meet with various lesbians and talk about their particular groups and the purposes of their organizations. Because I was a newcomer to the London lesbian scene, one woman who remained an important contact throughout the research volunteered to take me to a weekly meeting of a lesbian group and to introduce me to its organizer. My initiation into this lesbian group occurred in December 1974. This contact with a local lesbian group and its organizer proved to be important. It was at these weekly meetings that I soon became familiar with the London lesbian ghetto. I went along regularly to these meetings for a period of three years during the course of my research. It was through this particular lesbian organization, "Sappho," that I was able to distribute half of my questionnaires. I became known as a "resident sociologist'' and many women were willing to talk to me about my work. Thus in the first phase of her research, Betsy Ettore relied on participant observation, which in tum provided the foundation for preliminary survey research. She then shared her work with community members, a process many mainstream researchers consider very risky because community members generally do not agree with the researcher's view. On the other hand, Betsy Ettore describes this "sharing" as follows: The wealth of information which I gathered at these weekly meetings was invaluable. I was able to establish relationships of trust with many of the lesbians with whom I came in contact. Gradually, most members came to know me as a sociologist and as a confidant with whom one could discuss one's life. In order to build up relationships of mutual trust and understanding, I would periodically distribute my written work. Usually, my work was read with enthusiasm and often í was provided with pages of criticism, which proved useful in sharpening my own analysis of the lesbian ghetto. She also demonstrated her commitment to the women by her active involvement in the community. The number of settings that she observed and in which she took an active role was sizable: Along with these regular meetings I went regularly to bars, clubs and discos which were either all-lesbian or mixed gay (gay men and lesbians). Also, I attended various women's groups. The women's groups usually had a lesbian caucus which formed a working section of the organization. The groups, organizations or conferences of which I was a member numbered about fourteen. The trust she developed with the core group made it easier for her to do participant observation in related groups. From June 1975 until June 1976 I collected the major bulk of my research data. Since I had already become a trusted member of the lesbian community, my contact with other lesbians expanded into social contexts outside of my initial weekly meetings. Frequently, I was invited to lunches, dinners, parties, social gatherings. Also during this time I went to gay bars, gay clubs, lesbian bars, lesbian clubs, discos, etc. regularly. My amount of contact with the lesbian ghetto grew as my research progressed. A "promotion process" through the lesbian ghetto gave me acceptability in the ghetto, as well as validity in terms of my research role. It seemed to me that my analysis was becoming clearer and crystallized on a conceptual level. 206 Feminist Methods in Social Research After her extensive participant observation, Betsy Ettore was ready to draw on another method—survey research. As she writes: At this particular time, I wanted to test out my research concepts and in early 1976 I distributed 700 questionnaires. In February 1976 I attended a national Lesbian Conference. ... I distributed 400 questionnaires at this time; 101 were returned to me by post. A month later (March) I distributed 300 questionnaires to lesbian magazine subscribers who were affiliated either directly (actually attended some of the meetings) or marginally (knew about the meetings) with my Tuesday evening group, which had the same name as the lesbian magazine; 100 were returned to me. To complete the range of methods, she did two types of interviewing from March 1976 until December 1976: Interviews usually took place in people's home, my flat, place of employment, or at college. They lasted from between 30 minutes to two and a half hours. The average time was 45 minutes. I preceded my taped interviews with a discussion of what I was doing, the guaranteed confidence of the information and a general rundown of why I thought it was important for a sociological discussion of lesbianism to be developed. My formal interviews (20) were taped and followed a definite interview schedule. However, I often asked other lead questions which followed along with the main questions of the interview. My untaped interviews (40) usually centered around one or two lead questions, i.e., What is lesbianism? Why do you think you are a lesbian? In sum, Betsy Ettore's data analysis techniques combined ethnographic data analysis with quantitative devices that motivated her to develop a "computer programme, Lestudy, in order to facilitate this analysis." 19 Elissa Melamed's study of women's reactions to their aging20 also shows how different methods were useful at different stages of her research. In the methodological introduction she writes that her book Mirror, Mirror was sparked by an awakening consciousness that her identity had been tied to "pleasing and attracting others (particularly men) and relying on [her] youth and looks to do this." Having turned forty, she began to question who she was apart from that "charming but static young person." Her quest was for understanding and for methods with which to understand. I hungered for a perspective broader than my own personal history, for comrades in struggle, and for role models: women who had made peace with aging. What better way to get what I needed than to research the subject? I thought. My early explorations were tentative. I simply began talking to women, asking them how they felt about aging. These first investigations confused me. Most of the women I talked to deprecated the youth game and minimized the amount of time, money, and effort they spent playing it. It seemed as if everyone but the woman I happened to be talking to was buying the Oil of Olay and Loving Care. And then I realized that possibly other women felt as I had: ambivalent about aging, but afraid or ashamed to say so. Her initial research method—talking to women-puzzle she turned to self-disclosure: -was ineffective. To solve the Feminist Multiple Methods Research 207 I decided to confess to my interviewees that I had been having mixed feelings about aging myself that were hard to admit. The response to my candor was interesting. Immediately no one was neutral. Concern over aging was passionately denied (sincerely? defensively?) or it prompted outpourings of acknowledgment, confusion, resentment, and fear. . . . For me, this work of self-confrontation and sharing was profoundly healing. I realized that í had created my own personal consciousness-raising process. Communicating it to others and making it a mutual exchange seemed like the natural next step. This phase of her quest might be called "creating a focus group." One morning in 1978, sitting at home in the New Mexico sunshine, I decided to take that step—to create a consciousness-raising group for women over forty, a place where we could let all our age hangups show and, we hoped, re-envision the second half of life. . . . Confessing to our youth hangups. . . . And so we talked. . . . However, new obstacles appeared as our work continued. We became aware that the root causes of our problems were only partially in ourselves. No matter how ready we were to move forward in our lives, there wasn't that far to go. There simply were not enough well-paying jobs, interesting roles, or available partners for older women. In such an unsupportive environment, how could we grow to our full stature? As I began to probe this question, others presented themselves. Elissa Melamed continued her quest by reaching out for a cross-cultural perspective: I had begun my investigations in Santa Fe, a city of three cultures—Spanish, Indian, and "Anglo"—and had seen that aging was affected by culture. í wondered how women in other countries, not immersed in the youth culture of America, were dealing with aging. An opportunity to work in Paris gave me the chance to travel through Europe and continue my interviewing. This turned out to be easy. Since I was not attempting to do a formal study, I simply "followed my nose," letting it lead me toward a deeper understanding of women's lives. . . . I developed a questionnaire which I felt free to modify as the need arose; it is included in the appendix. ... í talked to academics, actresses, farm wives, prostitutes, jet-setters, housecleaners, lesbians, nuns, and others; I talked to over two hundred women in seven countries. . . . If I had ever thought that this was solely an American problem, that idea was quickly dispelled. ... It goes with urban living and the breakdown of traditional life styles more than with nationality. But it was nevertheless especially virulent among Americans. Even college-age women were affected (my youngest informant was twenty-one and my oldest, ninety-four). She then went on to include people who had studied the issue, and to include men. As the book took shape in my mind, I also felt the need to talk to various experts such as physicians, psychologists, and anthropologists. And I also questioned my male friends about their relation to aging. Finally, she wrote what so many feminist researchers write, as they focus on their population or social problem of interest: 208 Feminist Methods in Social Research The writing of this book had taken on a new urgency for me. ... I wanted us to create a context for making the public contribution that is as much our responsibility as it is our right. . . . Older women are currently our most underutilized natural resource—a resource we can no longer continue, to neglect.21 In sum, Elissa Melamed's methods began with self-examination, and then moved to interviewing, self-disclosure, action research (forming a consciousness-raising group), cross-cultural interviewing, and, finally, formulating a feminist agenda for social change. Responsiveness to the People Studied Some feminist researchers use a wide variety of methods because all of their respondents are important to them as individuals, and yet they cannot be studied uniformly. Researchers need multiple methods if they respect individual differences. An example of this set of circumstances can be found in a study by sociologist Athena Theodore, about academic women in protest. Athena Theodore allowed the people she studied to give her information in ways they considered most suitable for them. She collected standardized information only on "age, marital status, academic degree, field, rank or position, tenure status, and number of years employed at the time of their initial protest, or at the termination of employment, if a complaint or suit was instituted after leaving the campus." The primary material for her study was responses from approximately three-fourths (365) of the women in the final sample to an unstructured questionnaire. Imbedded in the questionnaire were items "suggesting a number of topical areas to help them recall experiences." The "topical reminders" in the unstructured questionnaire asked for an account of the full case, the internal grievance procedures used, encounters with government agencies and courts, interpersonal relationships relative to the case, women's studies and feminist research participation, experiences as graduate students, impact of the protest on lives and careers, and changes on campus relative to the protest actions. The amount of documentation was voluminous. Letters and memoranda, reports of various kinds, newspapers and newsletter accounts, government and court briefs and determinations were only a few of these. Some women sent entire files, their own taped biographies, and even scrapbooks containing the history of their cases. This research project also drew on contextual information, specifically "documentation . . . from other individuals and women's organizations having information about women with pending cases. . . . some of the women's campus colleagues, other academic friends, and spouses." No stone seems to have been unturned in her quest for information: Copies of documents from government agencies were solicited under the Freedom of Information Act. Combined with background data, contextual data, and unstructured questionnaires, forty of the women were also interviewed from periods ranging from 30 minutes to three hours, some on an intermittent basis, during the Feminist Multiple Methods Research 209 progress of their cases; approximately a dozen interviews were taped. Many telephone conversations, initiated by the women themselves occurred. Her responsiveness to the women and her commitment to integrate all possible forms of data led her to produce a particularly rich feminist study of the experiences of 470 academic women "fighting employment discrimination based on sex. To conduct her study of the in vitro fertilization experience in Australia, Ren-ate Klein combined questionnaires and interviews. She administered 40, 35-page questionnaires, followed by 25 interviews with women who responded to advertisements she had placed in leading newspapers in Melbourne in December 1986.23 Although her study was not as large or elaborate as Athena Theodore's, Renate Klein shared Athena Theodore's desire to go beyond the questionnaires to learn what specific women had experienced. This "going beyond" one method in-order to.leam about experience and its distribution is characteristic of much feminist research. Attempt to Apply Feminist Principles In a methodologically rich study of women's studies (WS) programs in the United States, the United Kingdom, and West Germany, Renate Klein combined questionnaires, interviews, and participant observation in a self-conscious attempt to apply principles of feminist research she had culled from relevant writing on this topic. She writes that "despite serious doubts about the validity of having people fill in questionnaires, [she] decided in favour of using a questionnaire as a starting point for later follow-up interviews." As she began distributing the questionnaire, she learned that respondents' willingness to complete it depended on whether they knew her. She then followed up with an interview of 37 of the 48 people who had filled out questionnaires in Britain. In addition, [she] conducted spontaneous interviews with 5 US, 12 British, 3 German, 3 Australian and 2 New Zealand WS teachers. . . . The interviews took place in University offices, restaurants, trains, parks, homes/student rooms, airports and even cars. The following is a description of her conversation-like interview process that included self-disclosure: In the cases where the interviewees had completed a questionnaire, we started our conversation on some specific remark they had made, but then attempted to cover the general outline of twenty interview questions which were also the basis for those who had not previously filled in the questionnaire. ... I was very pleased when the interviews turned out to be "conversations": two-way passages of exchanging information that not only benefitted the nature of the discussion by at times bringing up unexpected and exciting topics/insights/questions for both interviewee and interviewer, but also gave me the possibility of sharing some of my experiences/knowledge of WS with the women interviewed and thus relieved my sense of "exploiting them as research objects." 210 Feminist Methods in Social Research An additional data collecting procedure was participant observation on numerous campuses in England, the United States, Germany, Switzerland, and Australia. Just as some of her interviews had been serendipitous, so too was some of her participant observation: In the UK I got some of my most significant insights into the dynamics of Women's Studies—and areas of discrepancies between its theories and its practices— from conversations I had with the MA students on our weekly train rides from London to Canterbury and back.24 As is clear from these excerpts, and as she explicitly acknowledges, Renate Klein also utilized experiential analysis25 and cross-cultural analysis. Another example of applying feminist principles is discussed at length by Janice Radway in her study, Reading the Romance: Women, Patriarchy and Popular Culture. This project combined content analysis, observation, and interviewing methods to collect data concerning Harlequin Romance novels and the women who read them. She began her study with a critique of the standard methods of mass-culture studies that seek to recover "the meaning contained within a singular . . . literary text" and presume that readers imbibe that meaning in a uniform way. [S]ince I have abandoned the particular theoretical assumptions that would have justified the presentation of my own reading as a legitimate rendering of the meaning of the genre for those who usually read it, it was necessary to formulate another method for discovering the significance of the romantic narrative. Instead of this perspective, she adopted a stance called New Criticism that assumes readers construct the meaning of a text in relation to their own contexts. In accordance with its framework, she "developed the method . . . which depends heavily on questionnaire responses and on intensive interviews." Since she viewed readers as actively involved in constructing meaning, her method required that she study the readers, not just the texts they read. Nor could she presume that the texts compensate for problems in the readers' lives. Rather, she had to learn what the women's actual practices and readings consisted of. In her words: We are forced ... by the nature of meaning itself as the construct of a reader always already situated within an interpretive context, to conduct empirical research into the identities of real readers, into the nature of the assumption they bring to the texts, and into the character of the interpretations they produce. This methodological necessity led her to interviewing and ethnography: [My study] is grounded in an ethnographic examination of an actual community of romance readers who buy nearly all of their books from a single salesclerk named Dorothy Evans who has earned herself a local reputation as an "expert" on romantic fiction. My knowledge of Dot and her readers is based on roughly sixty hours of interviews conducted in June 1980 and February 1981. I have talked extensively with Dot about romances, reading and her advising activities as well as observed her interactions with her customers at the bookstore. I have also conducted both group and individual interviews with sixteen of her regular customers and admin- Feminist Multiple Methods Research 211 istered a lengthy questionnaire to forty-two of her women, most of whom are married, middle-class mothers. Janice Radway's notion of context extended beyond Dot and the readers, and entered into the economic matrix in which Romance novels are produced and marketed: [Rjomantic novels do not appear miraculously in Dot's hands or in those of her readers. They are, rather the end products of a much-mediated, highly complex, material and social process that involves writers, literary agents, publishing officials, and editors, as well as hundreds of other people who participate in the manufacture, distribution, and selling of books. Reading the Romance begins, then, with an investigation of this process and with a consideration of the possibility that recent changes in its organization and structure may well have contributed substantially to increasing romance sales. The assumption that readers could speak for themselves and had something to say led her to use a particular set of methods. These data-gathering techniques complemented her content analyses of the novels, a procedure she did not discard for the sake of the New Criticism. Respecting the women's views of their reading, Janice Radway also reports on her own interpretation shaped by her feminist consciousness. As she put it: I have attempted to offer an explanation of my informants' self-understanding that accounts also for motives and desires very likely felt by them but not admitted to consciousness precisely because they accept patriarchy as given, the natural organization of sex and gender.26 Her inclusion of the text, the talk, and the theory allowed her to show how the taken-for-granted assumptions of patriarchal society are part of the women's thought patterns and inform the way they interact. The Quest Image Feminist researchers who use multiple methods conjure up for me the traditional [male] image of the ' 'quest for truth.'' Feminists embarking on important research projects are like people setting out on important journeys. As the journey continues, they draw on different methods or tools.27 In their study of women academics, Nadya Aisenberg and Mona Harrington found that some researchers and scholars had actually internalized a self-image of being on a quest.28 Multimethod feminist research tends to be written in a way that reveals "the process of discovery." Initial discoveries energize the scholar to continue on her quest. When they fill her with excitement, she feels that she is on a passionate search. For example, Elinor Langer writes that in her study of Josephine Herbst, she acted "as if I were a traveler who had somehow gotten detached from my party, and Josephine Herbst were the rescuer sent out to bring me home." Being a researcher-traveler means having a self and a body. It means abandoning the voice of "disembodied objectivity" and locating oneself in time and space. As Elinor Langer wrote, "I am writing these words at a geographical place, 212 Feminist Methods in Social Research at a moment in time, at a point in my own history so far removed from the time I first [formulated the study]."29 It also means acknowledging that the self changes during the journey. Because of this experienced quest, many feminist researchers present then findings in a "process format." This format, in turn, is linked to their rejection of the discourse of positivism and objectivity. Psychologists Stephanie Riger and Margaret Gordon have written on this topic and suggest some alternatives to current practice: We perpetuate the myth that social research is free of bias by the use of the traditional scientific format for writing up studies. The introduction section of papers always discusses previous research, and rarely mentions a personal experience or observation of the researcher that stimulated the research. . . . The use of scientific protocol in report writing seems to be a ritual: if what we do looks like science, then it must be science. And the omission of the mention of human beings (except as subjects) implies that the research is untouched by human hands, and therefore unbiased. Recently some journals have made acceptable use of the first person singular "I" to refer to the researcher, instead of translation of action into the passive voice. ... In its own quiet way, this is change with revolutionary potential for social science, since it acknowledges that the research was done by a human being, with human limitations. Another small step with far-reaching implications would be to write the methods section of reports as a description of how the research was actually done, instead of reconstructing the logic of our techniques.30 A reflexive attitude31 toward the entire research process from "problem formulation" to "write-up" gives rise to psychological questions (why did I do this study?) and contextual issues (what were the interpersonal or structural effects of doing this project?). Stephanie Riger and Margaret Gordon provide an example: On our project on rape, two secretaries burst into tears while typing papers; in the course of trying to calm them, we learned for the first time that they had been rape victims. Publicity about our project in local newspapers stimulated telephone calls from rape victims who wanted to help us in our research, or who wanted our help. It is critical that researchers be responsible for the consequences of their research; yet many do not have the skills or inclination to be counselors. Supporting women's counseling centers, and referring victims to such services may be one solution to this problem.32 Feminist researchers who write about research in a "journey" format, as a process of discovery of which the product is a part, demystify discoveries. As projects proceed, new experiences are interwoven and new voices heard. The work process of the research becomes an integral component of the issues studied. The process becomes part of the product. This approach is humble since "findings" are housed in the project's specific features, rather than claimed as disembodied truth. Some writers begin the research report before the putative start of the project. They share with readers the childhood experiences that provide the context for undertaking the study in the first place. Sara Ruddick writes: Feminist Multiple Methods Research 213 Temperamentally I am a pluralist. From grade school, I welcomed the idea that there were many perspectives and hence many truths. In my childhood, it was Nazis, white supremacists, and later McCarthyites who claimed to speak from a privileged standpoint. Not surprisingly, when I studied philosophy I was drawn to traditions that rejected the ambition, pervasive among philosophers, of ordering ways of knowing from the last to the most adequate. ... In feminism too I have applauded those who reject the large picture for multiple perspectives.33 The person may describe how she came to write the book, why she chose the title, and even defects in the analysis of the study. Liz Stanley and Sue Wise's Breaking Out,34 Robin Lakoff and Rose Lynn Sherr's Face Value, and Chung Yuen Kay's At the Palace35 are vivid examples of studies where analysis of the author's background or experience augments more conventional research methods. In these books and others, the typical separation of the process of research from the product of research is eradicated. Concluding Thoughts When feminist researchers use multiple methods, the possible permutations and combinations are large. The particular combination of methods depends on the particular quest on which the researcher is embarked. Some of these quests are to link past and present; others to achieve heights of rigor, to integrate individual and social explanations of phenomena, and to test hypotheses generated in the field, as I have shown. The use of multiple methods requires a multiskilled feminist researcher or a researcher who is able to coordinate a team of individuals with a variety of skills. It requires time and resources. Many people are obviously willing to make the investment it takes to do this type of feminist research. Some cannot. There are compensations for the difficulties of acquiring multiple skills or making far-ranging commitments. One can be confident that a range of methods allows a range of individuals or circumstances to be understood in a responsive way. Important issues concerning women's lives can be understood in complex and thorough fashion. Researchers can then communicate this understanding to the public in a convincing manner. Multimethod research creates the opportunity to put texts or people in contexts, thus providing a richer and far more accurate interpretation. A review of multimethod feminist research examples shows that there are many reasons feminists adopt this approach. It is also striking to me, having seen how varied these research approaches are, that feminist research is driven by its subject matter, rather than by its methods. By this I mean that feminist research will use any method available and any cluster of methods needed to answer the questions it sets for itself.