We recruited 1,298 first-year students from vocational schools in the canton of Geneva, Switzerland. The participants had a mean age of 18.88 years, with the majority ranging between 15 and 26 years. Of the sample, 446 were female (34.7%) and 838 were male (64.6%). Most participants held Swiss nationality (n = 783; 60.3%), and nearly half reported French as their first language (n = 600; 46.2%). The majority (n = 1,032; 79.5%) were enrolled in a dual apprenticeship program, combining half-time school attendance with half-time training in a firm, while the remainder (n = 262; 20.2%) were engaged in full-time school-based apprenticeships.

The students recruited were grouped into 35 distinct fields of study, which corresponded to the occupational discipline that they were learning at the time we collected the data (e.g., business-related occupations, technical occupations, health and medical occupations, construction-related occupations; see Table 1)²2

The classification of fields of study was not an artificial grouping constructed retrospectively but reflected the categories of trades in which the students were being trained.

As further mentioned below, we also controlled for and included students’ endorsement of hegemonic masculinity as an individual-level variable.

We conducted the study in students’ classrooms during the course of a normal school day. They were all provided with a Qualtrics questionnaire and were asked to complete it individually on a computer. Total completion of the study took approximately 30 minutes and was supervised by the students’ regular teacher who was trained to ensure instructions were correctly understood and answer questions. All responses were fully anonymized, and participants were told so prior to starting the study. We obtained their consent before completing the study, as well as the parents’ consent of those under 18. We received ethical approval from an independent committee of the education research service of Geneva Canton, which was afterward validated by the education department of Geneva Canton. Note that this study was part of a larger longitudinal project designed to gather information and investigate determinants of gender-based and sexual orientation-based discrimination from students attending vocational schools in Geneva (Switzerland; for more details about the project, see Gianettoni et al., 2023). We certify that the analyses presented in this paper are entirely original, addressing aspects that do not overlap with any ongoing works related to this larger research project. Neither the study nor the current analyses were pre-registered.

As students (individual-level) were nested in fields of study (context-level), we performed a series of multi-level analyses and tested cross-level interactions between students’ sex and hegemonic masculinity across fields of study (i.e., estimated at the field level) on perceived and anticipated discrimination. To avoid generating context-level information by computing mean scores for each unit of the context (which does not correct for sampling and measurement errors and thus may lead to biased estimates; Croon & van Veldhoven, 2007), we used multi-level SEM and formed context-level constructs based on latent aggregations from multiple individual-level observed indicators (Marsh et al., 2009)⁴4

For information, we provided the individual-level mean scores of hegemonic masculinity by fields of study in Table 1.

First, we estimated empty models (intercept-only) and calculated the Intra Class Coefficients (ICC) and Design EFFects (DEFFs) to ensure that a substantial proportion of variance in perceived and anticipated discrimination can be attributed to the clustering across fields of study (k = 35). Second, we estimated baseline models by adding the main effects of sex and hegemonic masculinity (estimated at the context-level) on perceived and anticipated discrimination. Note that any positive effect of sex should be interpreted as a higher effect for female students on the outcome variable being tested (and vice versa for male students). Third, we expanded baseline models by adding cross-level interaction terms between sex and hegemonic masculinity (estimated at the context-level). Given that omitting random slope may inflate a Type 1 error when estimating cross-level interactions (Bell et al., 2019), we modeled a random coefficient model with the intercepts and the slope coefficients varying across fields of study. In the event of significant cross-level interactions, we performed decomposition analyses testing the effect of sex at 1 standard deviation above and below the cluster mean of hegemonic masculinity. We grand-mean centered hegemonic masculinity before being entered into the analyses and we used full-information maximum likelihood as an estimator. Because hegemonic masculinity as a context-level predictor is based on a latent aggregation of individual-level observed indicators whose referent is the individual and not the context (i.e., the field of study), we estimated, in both the baseline and cross-level interaction models, its effects at the field level after controlling for the students’ differences at the individual level (i.e., individual ratings of hegemonic masculinity; Marsh et al., 2012; Morin et al., 2014)⁵5

In line with Morin et al. (2014), who recommend not interpreting the effects of the individual-level ratings (notably because they may be confounded with the context-level effects), we did not report and interpret the effects of student-level endorsement of hegemonic masculinity on perceived and anticipated discrimination.

The present results raised several theoretical contributions. First, we demonstrated that female students’ discrimination experiences in certain fields of study may, in part, be related to the surrounding climate that students (but also teachers and any of those involved in the fields of study) put in place and reproduce through the type of ideologies they collectively adhere to within their field. In this way, we have added further evidence to the research showing that some educational domains are able to install and enforce particularly unfriendly learning environments for women, which tends, in turn, to have deplorable repercussions for their well-being, school satisfaction, and engagement (e.g., Casad et al., 2019; Settles et al., 2006). Our work specifically contributes to this body of research by demonstrating that an unfavorable field climate is associated with more gender discrimination. In essence, this suggests that experiences of discrimination do not emerge in a social vacuum but are crucially shaped by broader cultural contexts (Major et al., 2002), in which complex forms of ideologies are shared that create a fertile ground for the emergence of discriminatory acts, which afterward foster women’s perception of being discriminated against. Thus, discrimination does not primarily arise from hostile acts towards women, or from individual sexist attitudes, but from a socio-cultural environment in which members (both men and women) collectively share and spread a position that sanctifies the superiority of the male identity and justifies the use of disadvantageous treatment towards women. Gender-based discrimination practices develop more in socio-cultural spheres where ideologies promoting and justifying their application are collectively concentrated and disseminated (Brown & Bigler, 2004).

A key contribution of the current findings is to provide evidence that normative climate is associated with a greater incidence of discrimination experiences for women. In fact, our findings lead us to speculate on two routes through which normative climates may result in perceived discrimination. On the one hand, as we suggested above, one might argue that a normative climate of hostility toward women directly increases discriminatory behaviors toward women (especially because this ideological milieu offers an adequate basis for justifying such behaviors), which would inevitably increase perceptions of gender discrimination. Male students would feel authorized, and even encouraged, to treat female students unfairly and make stigmatizing comments, which progressively reinforce an unfriendly and high-pressure atmosphere over them. In turn, women may reasonably come to perceive that they are discriminated against and rejected. This fits in quite well with Kelly’s theorizing about the concept of a continuum of violence (Kelly, 1988), which argues that the forms of violence most suffered by women are also the most trivialized ones, that is, those that are defined as normal and ordinary in everyday interactions. It is the trivialization of masculine domination through the ordinary ideological climate, as something to be taken for granted, that makes sexist violence possible.

On the other hand, one may also conceive that a hostile normative climate could modulate how women apprehend social reality by shaping patterns of expectations. If a field normative climate leads to the expectation that women are disadvantaged and that men's dominant status is protected, female students can be more readily vigilant and inclined to detect discrimination (Major et al., 2002; Pinel, 1999, 2004). For example, Kaiser et al. (2006) showed that chronic or situational expectations of gender discrimination tend to direct women's attention toward subtle cues of threats to their social identity. In addition, gender expectations may shape the interpretation of personally negative events or treatments as acts of gender discrimination. Indeed, research has informed that those who can discern the presence of hostile views toward women are also more inclined to be sensitive to discriminatory treatment and interpret personally negative events or situations as an enactment of gender discrimination. Indeed, women differ in the extent to which they are cognizant of the stigma about women in society and those high in stigma consciousness report gender discrimination to a greater extent (Pinel, 1999), and make more attributions to discrimination (Pinel, 2004). For example, in a sample of women majoring in STEM fields, Casad et al. (2019) found that stigma consciousness increases sensitivity to gender-based rejection, thereby indicating that women’s propensity to perceive sexism in the surrounding climate is tightly related to greater odds of being sensitive to discrimination cues and perceiving oneself as a victim of discrimination. Thus, being aware of studying in a field whose ideological atmosphere is unfriendly to women or promotes traditional views on gender roles may lead to further interpreting negative treatment as a manifestation of gender discrimination (while they could see it as more harmless in other contexts), thereby leading to a higher likeliness of reporting it.

One of the strengths of our findings is the emphasis on the critical importance of field climate in relation, specifically, to hegemonic masculinity. Unlike other gender ideologies such as sexism, hegemonic masculinity provides an innovative and comprehensive framework for understanding gender power dynamics by focusing on masculine identity. Indeed, hegemonic masculinity, by prioritizing traditionally masculine traits and behaviors often at the expense of feminine ones, provides a broader perspective on the gender system, which encompasses a negative view of women as a subordinated category and, conversely, a positive view of masculinity as a dominant category. Moreover, it provides a unique lens for explaining the discrimination faced by both men and women. For instance, men who deviate from traditional masculine norms, such as those who are emotionally expressive or do not prioritize power, may experience social pressure to conform to these norms. By identifying the unique influence of hegemonic masculinity, we are better positioned to capture how gender norms serve to maintain gender hierarchies and how they are detrimental to both women and men.

From a practical viewpoint, our research has important implications. By pointing to the importance of normative climate in the emergence of experiences of discrimination, our findings encourage any instructors or advisors who wish to promote gender equality and discrimination reduction in the realm of education to engage in deep work on transforming the normative climate in the fields of study where hegemonic masculinity is the strongest. Insofar as we conceptualize normative climate as a system of beliefs shared in a broader, second-order context, it logically follows that any attempts aimed at its modification not only should focus on individual attitudes but on collectively shared representations of students and all actors involved in the field. We, therefore, suggest that collective interventions be implemented in schools (as well as in training firms in the case of dual apprenticeships) to deconstruct the norms of traditional masculinity, sometimes deeply rooted in the collective identity of some fields of study. It is fundamental that the beliefs or stereotypes that are part of the routine field environment can be revoked so that they are no longer embedded in the core socialization practices for newcomers or they no longer attract those who initially adhere to them and who target these learning environments in the hope of finding a familiar ideological ground.

There are several limitations of our research that are worth mentioning. First, our participants were students attending upper-secondary vocational schools. Given the particularities of vocational education (where schools are composed primarily of apprentices who wish to work quickly and who spend much of their time in training firms) and vocational education-related fields of study (which largely revolve around occupation types and not around disciplinary domains), it is important to remain cautious in generalizing our results to students of any type of education. Further studies could be conducted on samples of students from general education backgrounds (where gender discrimination is also highly prevalent across fields of study), at levels below or above upper-secondary schools.

Second, we operationalized normative climate using aggregations of individual-level attitudes from students nested in fields of study. Although this method remains common in establishing climate measures in education research (Marsh et al., 2012), alternative approaches may prove useful for future research. It might be interesting to create climate measures through objective indicators that pertain strictly to the field level and not to the student level, such as the number of recorded complaints for gender-related misbehaviors or possibly the number of interventions implemented in schools to prevent gender-based discrimination. In addition, the field climate is not only fed by the attitudes of students, but also by those of teachers, educators, and instructors. Therefore, future studies could consider refining climate measures by addressing the attitudes of all actors in the domain.

Third, we acknowledge that the number of fields of study included in our study (k = 35) was rather modest to meet the statistical requirements for adequately conducting multi-level analyses (which require a minimum of 50 to 100 context units; see Hox & Maas, 2001), which puts our analyses at risk of generating biased estimates. However, it is important to note that we could not reach more fields as the ones included cover all the existing types of fields in vocational education (which is divided into even more fields than general education) in the Canton of Geneva. To address this issue, we could focus on other levels of analysis, such as the class level. Nevertheless, we did not consider the class level because most students in vocational schools in the canton of Geneva do not spend more than two or three days in school and often attend courses that irregularly mix students from several classes.

Fourth, our study relies on correlational data that does not allow us to draw conclusions about causality and is based on a cross-sectional design that does not allow us to detect how changes in ideological climate over time might affect women's experiences of discrimination and whether, in turn, those experiences of discrimination might alter the shared ideological climate among students in the fields of study. A useful next step in addressing these concerns might thus be to resort to longitudinal designs.

In summary, our findings demonstrated that normative climate in fields of study is associated with students' discrimination experiences. The prevalence of discrimination among female students (compared to male students) appeared to be higher in the fields of study where forms of ideology are shared that support a culturally idealized view of masculine identity and the maintenance of a disadvantaged social status for women. From a broader viewpoint, these findings underscore the importance of considering the role of shared beliefs within local educational contexts to better understand students’ individual experiences, as well as their academic trajectories. Experiences of discrimination can be powerful psychological barriers to academic success and can put students at greater risk of leaving school. Our work suggests that the reproduction of gender disparities within fields of study, and more broadly within professional fields, derives, in part, from the persistence of shared traditional beliefs about gender roles and identities, which we urge all types of initiatives directed at promoting gender equality in education to primarily focus on.