Women, Gender, and Social Psychology

Virginia E. O'Leary, Rhoda K. Unger, Barbara S. Wallston, Virginia E. O'Leary, Rhoda K. Unger, Barbara S. Wallston

• Publisher: Psychology Press
• Year: 2014
• ISBN:
  9781317768555
• Edition: 1
• Topic: Psicologia
• Subtopic: Storia e teoria della psicologia

First published in 1985. This volume contains ten short pieces focusing on methodological issues, definitional problems, and new questions-both empirical and theoretical-that had been inspired by the social psychological study of women in the last decade. The material submitted was rich and varied, and offered an exciting' 'new look" at the current and potential contribution of the study of sex and gender to social psychology.

1

Integrating the Feminist Critique and the Crisis in Social Psychology: Another Look at Research Methods

Barbara Strudler Wallston

George Peabody College of Vanderbilt University

Kathleen E. Grady

University of Connecticut Health Center

The major issues raised by the study of women and gender (and social psychology) have been expounded independently but have much in common. The lack of cross-citation among these literatures is not surprising, given the narrow scope of most of our reading as well as the nature of the scientific enterprise with respect to citation (cf. Kuhn, 1970; Merton, 1968; Over, 1981). Thus, a number of discussions of problems in social psychology (e.g., Baumgardner, 1976; Buss, 1975; Sampson, 1977; Weissberg, 1976) stress the importance of the social context and its influence on the research process with no recognition of the parallel feminist critiques (e.g., Parlee, 1975; Sherif, 1979; Shields, 1975; Vaughter, 1976; Weisstein, 1970; Wittig, 1982). Engendering cross-fertilization for the benefit of methodology in social psychology and in the psychology of women is one of our goals in this chapter. However, we begin with a consideration of how the study of women and gender has contributed to the debate.

The study of women and gender has the potential for a dual effect on methods in social psychology. On the one hand, the numerous feminist critiques of sample selection, control groups, and interpretation of results can lead to a more careful application of existing methods. Thus, they can make our science more “scientific.” At the same time, there is a clear call for a broadening of available methods, more intensive descriptive research, more attention to understudied issues (such as rape, parenting, paid and family work), and an increased awareness of the political and cultural context for research.

The increasing emphasis on descriptive and correlational methods by researchers addressing issues of sex and gender has generated a debate about “hard” vs. “soft” methods that has often used the unfortunate labels of “masculine” vs. “feminine” (Bernard, 1973; Carlson, 1972). We believe that the gender of the researchers and the novelty of the questions asked are so badly confounded that it is impossible to assert that women researchers have affinity for particular methods. The scientifically sound use of observation and description to study relatively new questions with rigorous tests reserved for a more highly developed “state of the art” may better describe the current state of psychology research. In fact, some of our most distinguished colleagues have now, after a lifetime of rigorous empirical research, begun to employ more qualitative methods to address novel or intransigent questions (cf., Festinger, 1980; Schachter, 1982).

Although the potential for contribution in methodology is tremendous, the actual impact on the field of social psychology has been less than in more substantive areas. Sherif’s (1979) excellent discussion of bias in psychology and, in particular, the issue of status hierarchies in psychology suggests some reasons why this may be the case. Sherif notes that social psychology was at the bottom of the status hierarchy in 1943 and that the improved status of social psychology has come about through “self-conscious efforts to be accepted as experimental social psychologists. . . . The way to respectability in this scheme has been the appearance of rigor and scientific inquiry, bolstered by highly restricted notions of what science is about” (p. 98). We believe this continued emphasis on narrowly defined experimentalism has been one of the reasons that feminist methodological innovations and critiques have not had a substantial impact on social psychology.

In this chapter we have organized methodological issues into question formulation, sample selection, design, operationalization, the social psychology of the research process, statistics, and interpretation. These areas overlap extensively, and some of the distinctions, as well as the topic order, are clearly arbitrary. Over the months we have spent writing this chapter, everything written about methods or feminism has at some point seemed relevant. The resulting selection of citations has consequently been somewhat arbitrary.

One major goal has been the attempt to integrate the perspectives of the feminist critique and the “crisis” literature. As will become clear, it is our opinion that social psychology is getting older and better. Rather than simply rehashing criticisms, we try to highlight ways in which the study of women and gender has contributed to this process.

Question Formulation

How do we derive questions? Few of our research courses focus on the issue and our textbooks give little attention to question derivation. In general, we undervalue question generation (cf. Wallston, 1981), which is part of the art of scientific inquiry. McGuire (1973) has brought this to our attention as the first of seven koan “the sound of one hand clapping. . . . and the wrong hand” (p. 450). He notes that hypothesis testing has received 90% of our attention, while hypothesis generation has been neglected, “probably due to the suspicion that so complex a creative process as hypothesis formation is something that cannot be taught” (pp. 450-451).

In an unusual vein, and one that we applaud, Wrightsman and Deaux (1981) devote three pages of the methods chapter in their recent text revision to question formulation, and they note this in the preface as an important change. Unlike the traditional coverage of this issue, which assumes that questions are logical derivations from theory, Wrightsman and Deaux note theory as only one source of questions. They also give examples of observations of phenomena outside of the laboratory from which ideas have developed which are then translated into questions (and possibly theory) that are testable.

Frequently our questions in psychology reflect areas that are important and/or problematic to us. It is more than coincidence, for example, that much dual career research has been developed by members of dual career couples (e.g., Bryson, Bryson, Licht, & Licht, 1976; Rapoport & Rapoport, 1978; Wallston, Foster, & Berger, 1978). In social psychology, the emphasis on research on achievement and the nature of the definition of achievement certainly reflect the background and interest of the researchers (see Sutherland & Veroff, this volume). On the other hand, competence and achievement in parenting, homemaking, and volunteer activities have not been studied (cf. Wallston & O’Leary, 1981). This emphasis reflects the general valuing of traditionally male achievement in our culture (cf. Lenney, 1977; Mednick, Tangri, & Hoffman, 1975; Unger, 1979) and the socialization of researchers (female and male) into this value perspective.

Experience is a good source of questions. It may provide the best insights and most creative ideas. Unger (1981b) has noted that personal experience leads to questioning assumptions that others take as self-evident. Thus, as long as most psychologists are white, middle-class males, many important questions will remain outside the experience of most psychologists. Samelson (1978), in an historical analysis, has suggested that the change in psychological focus from the study of race differences to the study of racial prejudice could be accounted for by the entry of ethnic minority group members into the psychological profession. Unger (1981b) points out an analogous shift from the study of sex differences to the study of sexism. Thus, there are scientific, in addition to justice, arguments for changing the racist, classist, and sexist procedures by which students are selected and socialized. For example, Sherwood and Nataupsky (1968) have shown that the biographical characteristics of investigators were predictive of their conclusions regarding race differences in intelligence. Changing the faces of our student body is a first step. Allowing them to ask questions of interest to them, even or especially when those questions are beyond our experience, may be even more important in broadening our understanding of human behavior.

McGuire’s (1973) initial suggestions for hypothesis generation sources are worth reiteration: “case study, paradoxical incident, analogy, hypothetico deductive method, functional analysis, rules of thumb, conflicting results, accounting for exceptions, and straightening out complex relationships” (p. 451). Wallston (1981) explicates three additional principles for question generation: (1) experience which can be expanded by observation and discussion with others is a rich source of questions; (2) public policy issues (cf. Tangri & Strasburg, 1979) should be a focus of research; (3) situational as well as personological factors must be considered as potential causes of behavior when generating questions.

Methods for expanding our experience are critical to the formulation of good questions. Vaughter (1976) has argued for a participatory model in which research subjects and the public become an integral part of the scientific enterprise. The work of Plas and her colleagues (Plas & Bellet, 1983) to develop measures of Indian children’s value-attitude orientations exemplifies such an approach. The Anglo research team lived on an Indian reservation with Indian families in order to better understand the culture. Questions they developed for the scale were reviewed by elders of the tribe to check face validity for the relevant population. Although less extreme, the recent work by Azjen and Fishbein (1980) has stressed the importance of eliciting salient beliefs from a representative sample of the relevant population in order to evaluate attitudes. Ethnographers regularly validate their perceptions against those of research participants (cf. Corsaro, 1981, Leinhardt, 1978; Wallston, 1983).

Research on sex and gender particularly illustrates the failure to give sufficient consideration to situational factors (cf. Riger & Galligan, 1980; Wallston, 1978), although other areas (e.g., Caplan & Nelson, 1973) have shown this failure as well. Condry and Dyer (1976), for example, have presented an important reinterpretation of the fear of success literature from a situational rather than a motivational interpretation. Allen (1979) similarly notes that research on black women’s attainment has focused on personal characteristics and background rather than scrutinizing institutional and societal practices that systematically deny equal opportunity to black women. In fact, Caplan and Nelson (1973) found that only 16% of the studies of blacks included in six issues of the 1970 Psychological Abstracts examined situational causes rather than taking a personological approach (cf. Unger, 1981b). This tendency on the part of psychologists may reflect observers’ tendencies to overattribute the causes of women’s behavior to personal factors and men’s behavior to environmental factors (cf. Hansen & O’Leary, 1983; O’Leary & Hansen, this volume; Wallston & O’Leary, 1981). Not only have personal factors been overused to explicate female behavior, but sex (which is a personal factor) has been used as an explanatory variable, although it is frequently confounded with situational factors such as status or power (cf. Dion, this volume; Eagly & Wood, this volume; Henley, 1977; Miller & Zeitz, 1978; Piliavin & Unger, this volume; Unger, 1979).

To summarize, we have argued that questions and theory are undervalued in psychology. We have suggested some appropriate sources of good questions. Because a researcher’s experience is often a source of questions, it is important to broaden the experiential base of future researchers. The study of women and gender illustrates how the addition of female researchers has broadened the nature of the questions that are asked. Minority researchers and those from other social class and ethnic backgrounds would similarly enhance our understanding of human experience.

Design

We have already argued that the design should reflect the question. This is not to imply that there is a single design appropriate to each question (cf. Wallston, 1983). In fact, we can draw the best conclusions when several different methodologies are used to test the same hypothesis. Such triangulation (cf. Jick, 1979; Wallston, 1983) helps confirm that our findings are not tied to the specific methodology used. The important point is that some methodologies are not inherently better than others (cf. Labouvie, 1975), although fads in our field tend to suggest that they are. As Glass and Ellett (1980) note in their discussion of evaluation research, the best design is a compromise between the possibilities afforded by the situation and the research goal. Although choosing the situation allows some flexibility, practical constraints are involved in all research design.

Most of us have been socialized to think in terms of 2 x 2 analysis of variance models (cf. Rucci & Tweney, 1980). This limited design does not allow the investigation of complex processes when more than two variables with two levels of each may be important. Thus, our ideas regarding design, particularly to the extent that they are narrow, limit the kinds of questions we ask.

As social psychologists, we have been trained to value the quantitative over the qualitative, and experimental (manipulation) over correlational (measurement only) designs (cf. Wallston, 1983). Higbee, Millard, and Folkman (1980) found that 74% of journal articles in four mainline social psychology journals in 1978-79 utilized experimental methodology. There are stages of knowledge where qualitative and observational techniques may be particularly appropriate (cf. Depner, 1981), especially when we are investigating a new area and need to develop appropriate questions. Although such restrictions of the use of methods may be too limiting (cf. Trend, 1978), if we move too quickly toward manipulating one or two experimental variables, we run the risk of ignoring the most important variables because we have not sufficiently described the phenomenon of interest.

Taking advantage of powerful effects in the real world that cannot be simulated in the laboratory may necessitate giving up some control (cf. Cook & Campbell, 1979; Ellsworth, 1977). In most research we must consider the trade-off between internal and external validity (cf. Labouvie, 1975). Correlational studies in which the relationship between variables is investigated without assuming directional causality may be more appropriate for some questions (cf. Wallston, 1983).

Frieze’s (1979) important work on battered women demonstrates this problem. Investigating these important theoretical and applied issues is not possible using experimental methodology; thus, she cannot make definitive causal statements. However, this creative work also combines the more qualitative intensive interviews with structured questions to provide quantitative data. It is exemplary at fitting the method to the questions.

Wallston, Foster, and Berger’s (1978) utilization of experimental methods in combination with questions about the real-life experience of dual-career couples illustrates how different methods may complement each other. In a survey study, in addition to their personal job-seeking experience, respondents were provided with job-seeking situations and asked what they would do. The sex of hypothetical job seekers was experimentally manipulated through different forms of the situations. The hypotheses generated on the basis of responses to these situations provided a different approach to the actual experiences of these dual-career cou...