When an academic culture based on self-enhancement values undermines female students’ sense of belonging, self-efficacy, and academic choices

ABSTRACT

Although overall women are better represented in higher education than men, women’s psychological experience in various academic contexts is qualified by a decreased sense of belonging and academic self-efficacy, including in fields where they are not targeted by a negative stereotype. To clarify this phenomenon, we develop the hypothesis of a mismatch between female students’ values and the values associated with success in the increasingly selective realm of higher education. We argue that, whatever the fields of study, these values are self-enhancement values (competitiveness, self-affirmation, dominance). Three studies showed that when success was depicted in terms of self-enhancement values, women – but not men – expressed a lower sense of belonging, had lower self-efficacy and were less likely to pursue a given academic opportunity both in STEM and non-STEM fields of study. These effects did not appear in an academic context depicting success as being rooted in self-transcendence values (helpfulness, cooperation, benevolence).

KEYWORDS:

• Gendered values
• higher education
• sense of belonging
• academic self-efficacy
• academic choices

Data availability statement

The data described in this article are openly available in the Open Science Framework at https://osf.io/78vqc/

Open Scholarship

This article has earned the Center for Open science badges for Open Data and Open Materials through Open Practices Disclosure. The data and materials are openly accessible athttps://osf.io/78vqc/

Disclosure statement

No potential conflict of interest was reported by the authors.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1. An institutional ethics committee approved the research protocol. The research materials and the data sets are available on OSF: https://osf.io/78vqc/.

2. The results-pattern does not change when the three participants are included in the analyses.

3. The results-pattern does not change when controlling for the study setting.

4. A factor analysis conducted on the four items revealed a single factor with an eigenvalue greater than 1, accounting for 82% of the variance.

5. The results-pattern does change when we excluded the five participants who assessed the course as being intended more for men (responses 5 and 6 on the scale).

6. The results-pattern does not change when age was not entered into the model.

7. We also obtained a significant interaction between age and condition, B = −0.15, 95% CI [−0.27, −0.04], F(1, 223) = 6.54, p = .011, ɳ_p² = .03. In the self-enhancement condition, the older the participants, the lower their expected sense of belonging, B = −0.22, 95% CI [−0.41, −0.04], F(1, 223) = 5.61, p = .019, ɳ_p² = .03. In the self-transcendence condition, age was unrelated to sense of belonging (p = .267).

8. Although the interaction between age and condition was significant, B = −0.19, 95% CI [−0.35, −0.02], F(1, 223) = 5.14, p = .024, ɳ_p² = .02, the simple effects of age did not reach significance (p = .119 in the self-enhancement condition; p = .093 in the self-transcendence condition).

9. A significant interaction between age and condition was revealed, B = −0.54, 95% CI [−0.88, −0.19], F(1, 223) = 9.25, p = .003, ɳ_p² = .04. In the self-enhancement condition, the older the participants, the lower their interest in the course, B = −0.76, 95% CI [−1.31, −0.21], F(1, 223) = 7.36, p = .007, ɳ_p² = .03. In the self-transcendence condition, age was unrelated to likelihood of enrolling (p = .146).

10. It is important to note that the RMSEA tends to be substantially elevated for correctly specified models with small degree of freedom (Kenny, Kaniskan, & Mccoach, 2015). In fact, the authors recommend completely avoiding computing the RMSEA when model dfs are small, and particularly when the χ² test is not statistically significant, since, in such a case, we know that the model relatively closely reproduces the data.

11. Given that in the alternative model we did not have a specific prediction about the degree of mediation for the effect of condition on sense of belonging through academic self-efficacy, we also ran Model 2b, in which we included a direct path from condition to sense of belonging. Model 2b also showed relatively poor fit to the data, χ²(1) = 2.98, p = .084.

Additional information

Notes on contributors

Cristina Aelenei

Cristina Aelenei is Assistant Professor in the Social Psychology Department at Paris Descartes University. Her research examines how ideological and structural factors (e.g., cultural mismatch, backlash effects, glass-cliff dynamics) contribute to the maintenance of existing social hierarchies.

Delphine Martinot

Delphine Martinot is Professor of Social Psychology at Clermont Auvergne University. She is particularly interested in the effects of academic comparison contexts on school disengagement, self-esteem, and endorsement of system-justifying beliefs.

Alyson Sicard

Alyson Sicard is a PhD student at Clermont Auvergne University. Her research interests include gender inequalities and the consequences of threatening school contexts.

Céline Darnon

Céline Darnon is Professor of Social Psychology at Clermont Auvergne University. In her research, CD examines how the two functions fulfilled by the educational system, education and selection, affect the goals students report, their social value and contribute to reproduce and legitimate class and gender inequalities.