Introduction

Qualitative methods predominate in Brazilian educational research, and the same is true for studies that investigate the role of gender in educational processes. Through some examples of research, this article discusses the main contributions of qualitative methods to this area and points out the limits of these approaches, especially the lack of effective dialogue between qualitative and quantitative research.

Research on gender relations is an area of educational studies that has been expanding significantly in Brazil since the turn of the century. In 2001, Rosemberg (2001) produced a systematic survey of Brazilian academic production on education and gender, indicating that the number of studies on the subject was small but growing. Until the late 1990s, the issues addressed were scattered and dialogued little with the central issues of the educational field, particularly formal education. The 2000s witnessed a significant growth of production in this subfield, in its theoretical density, and in its approach to the burning debates in school education, as Vianna and collaborators (2011) indicate. Analyzing theses, dissertations, and articles on gender, sexuality, and formal education in Brazil published between 1990 and 2006, the authors found that in this period production on the subject grew more than the total of researches in the educational area as a whole.

Today, gender research is represented at the main scientific congresses of the educational field, as well as in the reference journals and graduate programs in education. Mainly through empirical research, it approaches themes ranging from teaching to sex education; from homophobia in school to differences in school performance between the sexes. And it ranges from day-to-day day care to higher education and academic careers, as well as educational policies, informal learning, and cultural products.

Qualitative methodologies predominate in these investigations, with methodological tools including observations, open or semi-structured interviews, life stories accounts, and discussion groups. Field diaries as well as audio and video recordings are used as recording tools, enabling researchers to transcribe and register their findings.

Breaking with Naturalization

These methodologies have been particularly important in uncovering the subtle forms of gender inequality in everyday life. One of the main mechanisms for perpetuating these inequalities is their naturalization, a mechanism that presents social relations as immutable facts of nature, derived from the corporal differences between the sexes. Stated as evidence, embedded in the order of things, inequality between the sexes appears as normal and inevitable, and even organizes our way of thinking about, classifying, and understanding the natural world and social practices. As Bourdieu pointed out (1995, p. 133) in a study of relations of domination between the sexes, the analyst deals “with an institution that has been inscribed for millennia in the objectivity of social structures and in the subjectivity of mental structures” and “tends to employ as instruments of knowledge those categories of perception and thought that should be treated as objects of knowledge.”

Epistemological shifts connected with the effort to “make strange what is familiar,” so common in ethnographies, are useful in overcoming these pitfalls, and Brazilian research on gender in education has used this approach with good results.

A good example of how the careful use of ethnography can produce positive results is a study by Jung (2003) carried out in an elementary school in a small community founded by German settlers in southern Brazil. The author started the research interested in exploring the ethnic-linguistic implications in the process of school literacy in Portuguese for children who only spoke German and those who were German–Portuguese bilinguals. But during field research she came across the salient issue of gender: in the classroom, clues pointed more strongly toward differences in learning for boys and girls—with better outcomes for girls—than ethnic differences exemplified by language. Thus, Jung reformulated the research problem and proceeded from the hypothesis that the way boys and girls experienced German ethnic belonging, as well as literacy in Portuguese, was associated with their gender identity. She accompanied the students not only in school but also in different spaces—church, home, farmers’ association—making observations and recording video episodes. She concluded that the men in the community affirmed a local identity based on the German language and manual labor in the field, in accordance with the identity of German settlers and in contrast to the men of neighboring urban areas. Meanwhile the girls and the women, whose tasks did not involve work in the field, which was deemed masculine, embraced an urban, Brazilian, Portuguese-language identity, and were more oriented toward neighboring areas where they saw better opportunities for social ascension. This search meant an effort to obtain the prestige that they did not have within their community of origin and it represented a valorization of school and Portuguese literacy. At the same time, the school recognized and valued a feminine identity associated with characteristics such as organization, cleanliness, and discipline, relating disorganization and lack of bodily cleanliness—associated with the male identity of the German settlers, to the boys. The author concluded that the German language was a trace of local identity that women sought to erase, since rural life was an unattractive choice to them; and also that this dynamic contributed to girls’ success in school.

It is clear that Jung’s results were only possible due to meticulous ethnographic work, based on solid theoretical references, which allowed the research to grasp the gendered meanings prevailing in that community. These were not obvious and could easily be overlooked in an investigation that was not open to the surprises of fieldwork or that only used a methodology incapable of creating distance between the researcher’s assumptions and local notions about gender relations. Ethnography offers powerful tools for creating doubt given its emphasis on the difference between the symbolic universes of the researcher and the researched, highlighting “the gaps and asymmetries between our way of seeing things and that of others” (Fonseca, 1998, p. 59).

In addition qualitative methodologies can contribute to face two other challenges present in research on gender and education and highlighted by Jung’s (2003) study: the importance of articulating gender inequalities with other social structures of domination, such as class, race, and ethnicity; and the need to complicate male–female relationships to understand girls’ school success. In other words, educational research is challenged to look for elements that allow us to understand how girls have been achieving better school results (in Brazil and in most Latin American countries, as well as in countries of the Global North) given women’s subordination in other social spheres, such as the labor market, politics, and family life.

Avoiding Essentialisms

The success of girls in school has been the central theme of my research, with a focus on elementary schools (ages 6 to 14) serving working-class urban areas. Along with graduate students, I examined how teachers evaluate boys and girls and how racial inequities interfere in this process; how children themselves perceive and participate in these evaluation dynamics; and how family socialization influences how boys and girls perform in school (Carvalho, 2004, 2005, 2009, 2015; Pereira & Carvalho, 2009; Senkevics & Carvalho, 2015).

In almost 20 years of research, using qualitative methodologies was of paramount importance to avoid explanations based on universal notions of women, men, femininity, and masculinity.

At every stage of the research, we sought to limit our findings to available generalizations. There are few studies in Brazil about the gender-differentiated school trajectories of indigenous peoples (Teixeira & Gomes, 2012), or among non-indigenous people in rural areas, as well as among urban middle-class families (Glória, 2007) or elites. But we can suppose that the same phenomenon of girls’ school success, verified transversally across society, results in each social group from processes different from those we observe among the urban working class.

Among the Xakriabá Brazilian natives of northern Minas Gerais, for example, Teixeira and Gomes (2012) found a higher female presence in the final grades of elementary and high school, and in this last stage there were three times more girls than boys. In anthropological research based on prolonged visits to Xakriabá indigenous land, in which observations and interviews were made, the authors related this situation to the changes taking place in that social group, among them the monetarization of relations with the consequent expansion of paid work; the increasing presence of women in public jobs (mainly as teachers and health workers); and their greater participation in public decision-making spaces, such as village assemblies. Teixeira and Gomes reiterate that “despite the points of convergence between the Xakriabá indigenous school and the Brazilian non-indigenous school regarding the better progress of women in school … different meanings are attributed to the process of schooling and literacy because they settle in contexts with specific social, cultural and economic dynamics” (2012, p. 55).

On the other hand, in analyses also developed in other countries, we find explanations for the school success of girls that are very similar to ours. For example, most of the girls we studied, who lived in slums in the metropolitan area of São Paulo, had much better defined future dreams than their male siblings and classmates. While the latter spoke vaguely about being soccer players, astronauts, or firefighters, the girls wanted to be doctors, actresses, veterinarians, biologists (Carvalho, 2015; Senkevics & Carvalho, 2015). These girls lived a virtuous circle: successful and comfortable at school, they dreamed of qualified professions that demanded long schooling, and soon established ambitious aspirations. The existence of these dreams, in turn, could be driving them to invest more in school, value learning, and achieve good school results.

Similarly, Ames (2013) describes how indigenous girls in rural Peruvian areas had high educational aspirations and, through education, sought to break with both poverty and gender oppression, just like the girls we studied in São Paulo. Through research based on observations and interviews in three different rural locations in Peru, the author studied girls between the ages of 12 and 13 and showed that for them, extending their time in school implied distancing from traditional conceptions of femininity that imposed upon women obstacles such as early marriage, pregnancy, and the life of a homemaker as priorities over their professional life. Girls’ future plans, therefore, not only considered careers hitherto little experienced by the other women in their social group, but also involved a revision of the gender representations that guided their notions of femininity, trying, ultimately, to escape the limiting experience of their mothers.

The reference to these different studies on girls’ school success is to point out that, from the methodological point of view, these comparisons are only possible, even when they point to similarities, by dense reference to the specific contexts of each of the studied groups of girls, not from a universal conception of femininity. This universal conception underlies essentialist interpretations, very common in gender studies at the international level, and explains “women’s” school success through supposed passivity and submission to school norms.

Linking Different Social Determinants

In our research, we also tried to understand the concept of gender in relation to other social structures, such as class and race—an effort known within feminist studies as “intersectionality” (Collins, 1991; Crenshaw, 1991). This has led us to restrict the conclusions “to girls from urban popular sectors” or “to black boys from low-income families,” for example. With this kind of care, it was possible to perceive that it is not the “school failure of the boys,” but of the black boys of the working classes. Broad arguments about girls succeeding at school because they are obedient did not explain different school trajectories among girls. Or that not all schools, and not all the teachers of the same school, valued organization and discipline over autonomy and creativity.

Thus, within the investigated group, we used qualitative methodologies not only to perceive the differences of meaning between being a boy and being a girl, but also black and white, and a higher versus a lower family income. Linking these social inequalities allowed us to detail everyday processes in the classroom and to better understand the daily construction of different school trajectories for white and black girls, white and black boys, belonging to different social strata.

To explain more clearly how qualitative methodology allowed us to advance in the understanding of the complex processes of articulation between different social inequalities, it is worth noting a specific investigation by way of example. Between 2002 and 2003, I researched school evaluation processes in the initial grades of primary education (6 to 10 years of age) in a public school that cared for children from both low-income and middle-income families (Carvalho, 2005). Among the results, I specifically discussed the differences between the racial classification of the students by the teachers and by themselves (self-attribution), trying to show the extent to which the classification made by the educators was related to children’s school performance.

In Brazil, racial classification relies on both appearance (phenotypic characteristics, skin color, or hair type) and ancestry, as well as socioeconomic status of the person, and it is also influenced by schooling and sex. It is a fluid and variable racial classification, with the possibility of exceeding the color line due to the combination of appearance and social status. Investigations need to deal with the complexity of the racial classification process thus constituted, in which phenotypic characteristics, imbued with value and charged with meaning, make up a system operated in a context of social interaction (Araujo, 1987).

This focus emphasizes the production of meanings associated with phenotypic characteristics and color assignments, as well as the interrelationships through which these classifications of color and race emerge. Racial identification is perceived as a lifelong social process that does not stem from an immediate perception of natural data. From this perspective, racial identity is dynamic and must be taken not as an objective and unquestionable fact but as a changing system subject to social perception, in which various factors unrelated to physical appearance carry considerable weight. It is necessary to consider both the self-classification of each subject, as well as the classification by others, in addition to the relational context in which these attributions of color or race take place.

In Brazilian society, alongside the income of the interviewee, there are indications that education, housing, as well as characteristics of the interviewer such as race, schooling, and sex all have a significant influence on the subjects’ racial classification (Piza & Rosemberg, 2002; Telles, 2003).

It is easy to see that qualitative methodologies are particularly important to understand these processes, since we are dealing with the social construction of meanings and the contexts of social interaction.

In the case of the aforementioned research (Carvalho, 2005), I tried to understand if the teachers’ color classification of the children varied according to their gender, family income, and school performance, comparing the teachers’ classification with that of the children themselves. I did not regard any of them as the true or proper classification, and the main purpose was to understand their meanings. The initial hypothesis was that, within the school, the teachers’ racial classification would refer to not only phenotypic characteristics, gender, and socioeconomic level—elements present in Brazilian society as a whole, but also to the student’s school performance. I dealt with 240 children who were taught by eight teachers in classes of 30 students each, which led me to operate with numerical pictures and percentages, as well as observed situations and recorded interviews. The research maintained its qualitative perspective, since there was no pretense of constructing samples, obtaining statistically representative results, or establishing probabilities.

In general terms, teachers tended to whiten their students, comparing their classification with that made by the children themselves. But there was a tendency for the teachers to make the girls whiter: while 60% of the girls perceived themselves as black, for the teachers only 26% fit that description. For boys, the rates were 55% and 30%, respectively. In general, income also had a marked influence: the comparison between the teachers’ classification and the self-classification of the students in each income bracket shows that, within the broad propensity to whiten students, they tended to further decrease the proportion of blacks among high-income people and, conversely, to increase that proportion among low-income people.

In addition, the analysis of the empirical material led to the suggestion that, in school, the teacher’s classification of race would be influenced by the existence or absence of school problems, in particular learning difficulties, with a clear link between belonging to the black race, masculinity, and difficulties in school. No results from the children’s performance in standardized tests were used since the aim of the research was not to evaluate learning but to understand the trajectories of the students in the school. These trajectories are defined by the teachers’ evaluation, which determines if the child will require tutoring, will be retained at the end of the school year, and so forth.

The results allow us to affirm that the attribution of race made by the teachers was not exclusively related to the phenotypic characteristics of the children, to their gender, or to the perception they had about the income of their families. It was also tied to their school performance. As the evaluation used in this case was constructed by the teachers themselves, we can assume that they tended to perceive children with learning problems as black, and also tended to evaluate negatively or more rigorously the performance of children perceived as black, a dynamic that was particularly intense in relation to boys.

The conclusion was that there was a clear association, in the frame of reference used by the teachers to evaluate the children, between a type of black masculinity and low performance in learning, which certainly had consequences both for the school trajectories of black boys and for the construction of their racial identities. On the other hand, the girls classified as black by the teachers were the least cited by them in the interviews, either for praise or to report problems, a picture of invisibility.

I have no doubt that the use of qualitative methodology was key to highlighting these intersections between socioeconomic level, gender, and racial belonging in the daily life of the school. The challenge of constructing intersectionality, linking gender inequalities to other social structures, finds in qualitative research methods a powerful ally because it allows one to understand how each form of inequality combines with the other, creating new meanings. This leads to analyses beyond the summation of subordinations—double or triple discrimination—already widely criticized by feminist literature (Nicholson, 1994; Scott, 1988). It is not that within the group of boys some are also black and others are still poor. Nor is it that, from the common condition of girl, some are differentiated by being black and/or poor. It is about understanding the meanings of being a high-income black boy, or a poor white girl, for example, each of these combinations resulting in different social places with different expectations and possibilities.

The Actions of Subjects and Social Structures

Another revealing example of the potential of qualitative methodologies for the study of gender relations in the educational field is Cavaleiro’s (2009) research on homosexual girls in the school environment. From observations in a high school, holding discussion groups, and recording semi-structured interviews with teachers, employees, and seven girls aged 16 to 17 years, Cavaleiro presents a very detailed and nuanced picture. Most Brazilian research on homosexuality at school confines itself to denouncing the homophobia and suffering faced by homosexual students and educators. Without denying the importance of this denunciation, Cavaleiro’s (2009) study went on to reveal the forms of resistance and the methods invented by non-heterosexual students to live their identities, their sociability, and their sexuality. The study was able to indicate the uses of school times and spaces in an alternative way, tactics and skills present in everyday life that allowed the girls to break through surveillance and tighten controls, as well as to create alliances and solidarities. The researcher describes this dynamic as a game of forces and shows the use of the ladies’ bathroom and little visible corners of the patio for dating; times stolen from class; the tension experienced by the girls between revealing and hiding their identities and practices; and the choice of friendships and partnerships, be it among students or among educators.

Thus, this research foregrounded one of the contributions of the qualitative methodologies that seems to me most relevant: the possibility of making visible the action of the subjects, revealing that all power relations include resistances, tensions, and ruptures. It was the ethnographic view that allowed the researcher to go beyond denunciation and present the girls as active subjects in the construction of their lives and not as mere victims of prejudice and discrimination, reversing a tendency present in gender studies and education, especially those about non-heteronormative sexualities.

Furthermore, Cavaleiro (2009) points out that the control mechanisms, epitomized by the requirement that homosexual girls be “discreet,” served to legitimize inequalities of status within the social structure. “Ultimately, therefore, we are talking about social inequalities” (p. 180). Thus, it is a good example that highlighting the subjectivities, the changing conditions, and the subject’s actions is not contradictory to the perception of gender as a social structure.

In this way, it seems to me that Cavaleiro’s (2009) work is also a good example of analysis that goes beyond the specific situation observed, of the “cases” under study, a move not always seen in research using qualitative methods. The emphasis of contemporary Western societies on the individual reverberates in scientific thought and often leads analyses to isolate the individual and her/his subjectivity from her/his social group. Particularly in gender studies, the focus on the individual construction of identities, a common approach in the literature of the countries of the Global North, especially in the United States, often obscures sociological analysis from the inequalities, asymmetries, and hierarchies stemming from the social relations of gender.

In everyday life in Brazilian cities, talking about gender has become synonymous with talking about individual orientations regarding body, sexuality, and personality traits relating to femininities or masculinities, as if these were individual choices outside a social and historical context. Moreover, even in national academic research, talking about gender as a social structure requires explanation, for one immediately finds criticism of determinism. In this context, qualitative methodologies, when studying few subjects and focusing attention on the solutions that each individual finds in her/his trajectory, often give way to approaches that are not attentive to social conditions.

That is why it is important to recall Connell (2015), paraphrasing Marx: “We make our own gender, but we are not free to make it as we please” (p. 156). Researchers who use qualitative methodologies have the theoretical commitment to reinstate their analysis of particular cases within broader political and social relations, emphasizing that gender is a social structure.

Reflecting on the uses of ethnographic research in the Brazilian educational field, anthropologist Claudia Fonseca (1998) criticizes strictly subjective and individual approaches, noting that “particular data opens the way for broader interpretations” (p. 60). In order to properly interpret the behavior he/she observes, the researcher must necessarily draw from social life as a whole, its material and symbolic patterns, in order to place specific situations and individuals within their social totality. This is not a novelty, but rather a part of anthropological tradition. For example, in his introduction to Mauss’s work, Levi-Strauss reiterates that individual behaviors are never symbolic in themselves, but “are elements from which a symbolic system, which can only be collective, is constructed” (1974 [1950], p. 13).

This is not to say that the subjects are totally predictable and are bound by social determinations or that there is no prospect of change. The concept of structure serves to capture the fact that there are patterns that are powerfully determined in relationships, patterns that are long-lasting. If the gender arrangements of a society are a social structure, they do not mechanically define how people or groups act. Rather, they define the horizons, terms, possibilities, and consequences of that action (Connell, 2015). Structures, whether material (as in Marxist structuralist studies) or symbolic (as in the vast production on gender affiliated with Cultural Studies), are not separate from everyday life and are constantly updated by human practice.

Because of its potential for analyzing everyday life, meaning construction, and social interactions, qualitative research can contribute decisively to the understanding of the simultaneity between the actions of the subjects and social determination, between change and permanence, between individual and social. However, for this, in addition to maintaining the inseparability between material and symbolic dimensions, the researcher must fully accomplish the link between the particular and the general.

A Dialogue to Be Developed

In this effort, it may be useful to also break a barrier that is frequently found in Brazilian educational research—that between qualitative and quantitative studies. By dealing with macro-social data, research with quantitative methods allows for a more direct focus on social structures. It is not a question of establishing an opposition here between these ways of approaching the objects of research, supposing some kind of epistemological incompatibility between them. Neither is it intended to develop the necessary theoretical debate about the limits and potential of quantitative research tools and their complex relationships with qualitative research.¹ The idea is only to point out this gap and try to perceive some practical limitations that have hampered this dialogue.

In Brazil, most research in the field of education results from theses or dissertations. Few studies are developed by teams, in addition to being usually resource-poor. All of this hinders the production of information on a large scale through quantitative research or combining qualitative and quantitative research. With regard to official data, government bodies collected little data on the quality of school education and demographic information on racial belonging until the 2000s. In addition, it was not always possible to access detailed data.

For example, in the case of the School Census, an annual survey of national educational data carried out by the National Institute of Educational Studies and Research Anísio Teixeira (Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira, or INEP), data on color/race only started to be collected in 2005. And it was only in 2006 that the process as a whole ceased to be on paper, with the creation of an online system that collects individualized information from students and teachers (Senkevics, Machado, & Oliveira, 2016).

Today, education researchers can easily access demographic data from the Brazilian Institute of Geography and Statistics (Instituto Brasileiro de Geografia e Estatística, or IBGE), among which I would highlight the National Household Research Sample (Pesquisa Nacional por Amostra de Domicílios, or PNAD), which allows for the annual monitoring of data. Researchers can also see individualized data from the INEP database, the agency responsible for the Basic Education Census, with annual information on all enrollments, schools, classrooms, and classroom professionals; as well as the Census of Higher Education, which collects annual information on enrollments, courses, teachers, and institutions of higher education.

Access to this information has been facilitated, and local or regional data can also be obtained directly from government sites, sorted by gender, and with the ability to cross-reference them with different variables such as race or income, which allows for studies with fewer resources and smaller research teams.

However, the training in statistical analysis offered to researchers in education—and, among them, those interested in gender relations—is precarious. As a result, most large-scale data analyses are still produced by economists or other professionals with no background in education. Therefore, education researchers’ choice of qualitative methodologies does not always stem from how well suited they are for a given research problem, but rather from the researcher’s limitations or restrictions on available data. As a consequence, qualitative research predominates in the educational field and there is little dialogue between investigations that use different methodologies.

Good examples of Brazilian quantitative research on gender and education are the studies by Alceu Ferraro (2009, 2010, 2012b), which mainly use microdata from the demographic census; Fulvia Rosemberg and Nina Madsen (2011), based on tabulations of the National Survey of Household Sampling (PNAD) and the Census of Higher Education; Artes (2007), based on the National Indicator of Functional Literacy (INAF), a non-governmental index composed of a characterization and test questionnaire; and Artes and Carvalho (2010), which uses data from the PNAD.

On the other hand, there is practically no research on gender and education that combines quantitative and qualitative approaches. For the sake of example, I present some questions that would benefit from this interaction between different research tools. One can cite questions arising from the recent expansion of access to secondary education in the country: what does large-scale data say about the access and permanence of girls and boys in this stage of education and how do they differ by race and income? With these data in hand, qualitative studies might address the meanings that girls and boys of different backgrounds and racial identifications attribute to secondary education, what their future plans are, and how they articulate their entry into the labor market and continuing education.

Questions about the expansion of early childhood education might also be explored: how much has the enrollment of children up to six years of age grown? Where are the day care centers? How are the families of children cared for? How many teachers were hired to promote this expansion? What are the sociodemographic characteristics, the type of hiring, and the training of these new early childhood educators? From this information, qualitative studies could investigate the resulting changes in the lives of mothers whose children had access to kindergartens and preschools, including changes to ideas of childhood and maternity, as well as in the family structure and paid work of women. One could also understand the impact of the new hiring of female teachers on the lives of these women, now employed in an occupation strongly associated with traditional femininity.

Conversely, qualitative studies on the association between meanings of black masculinity and school difficulties in the daily practices of teachers are asking for quantitative investigations that can evaluate the extent of this process and its consequences. What data are available on the trajectories of schooling for black boys? Would considering the results of standardized tests instead of the evaluations made by teachers yield different results?

These are just a few possible examples of how gender and education studies could develop in their specific field from dialogues between these various approaches, opening space for issues highlighted in statistical analysis to be explored in qualitative research, which in turn might generate new questions to be investigated in macro-social databases.

This article has noted that qualitative research methodologies contribute to gender and education studies in Brazil as powerful tools that break the naturalization of gender relations and in the construction of contextual studies that avoid essentialist statements about men and women. The article indicates that researchers can use qualitative methodologies to link gender to other social determinations in an “intersectional” perspective; and that analysis from qualitative data may help break the dichotomies between social structures and individual action. Finally, the conclusion draws attention to the need for greater dialogue between quantitative and qualitative research in the area of gender and education studies.