Participation in higher education is highly stratified by socio-economic status and gender. University enrolment as such is strongly determined by the socio-economic backgrounds of the parents (Barone et al., 2017), while gender differences mainly occur with respect to the field of study. It is a widely acknowledged finding that women engage less in Science, Technology, Engineering and Mathematics (STEM) fields even when math-related ability is controlled for (Zafar, 2013; Reuben et al., 2017). Both phenomena contribute to the reproduction of social and gender inequalities, but also exacerbate a lack of skilled workforce in the labour market in general and STEM-related occupations in particular. Correspondingly, there is a growing number of educational interventions which aim at fostering college enrolment and STEM choice of females and pupils from disadvantaged backgrounds. The vast majorities of these interventions focusses on reducing information barriers, i.e. by providing information about costs and benefits of college enrolment (Barone et al., 2017; Barone et al., 2018; Ehlert et al., 2017), as well as assistance for the application process (Oreopoulos and Dunn, 2013; Bettinger et al., 2012).

While tackling information deficits has indeed turned out to be an effective means of reducing inequalities in college enrolment and field of study choice, we offer a novel approach to alleviate gender and SES segregation in higher education, which highlights the role of psychological resources. Since educational research has convincingly shown that psychological resources play a major role in shaping gender and SES differences in the career choice process of adolescents (Perez-Felkner et al., 2015; Wang et al. 2014), targeting these resources will be a promising way of addressing stratification in participation in higher education. For example, research has shown that girls hold lower perceptions of their individual abilities in mathematics, which in turn influences their STEM field of study choice (Correl, 2001). Given that STEM occupations are perceived as requiring strong intellectual ability and talent to succeed (Leslie et al., 2015), girls might need particular reinforcement to consider these fields as a viable option. The contribution of social-psychological interventions to address inequalities in behavioural outcomes at the transition into higher education has received limited attention so far, although a vast amount of literature documents its effectiveness in reducing achievement gaps and stereotype threat in the school context (Yeager and Walton, 2011).

In this paper, we therefore analyse the impact of a psychological counselling approach for secondary school pupils on psychological resources as well as field of study aspirations. The approach consists of two components that shall foster a) students’ self-efficacy beliefs and b) students’ growth mindset. First, participants meet an already enrolled student who shares his or her experiences on real problems as well as possible solutions during his or her studies. This is supposed to increase the consciousness about available resources in case of problems, hereby leading to greater self-efficacy. The second part closely mirrors a growth mindset intervention that aims at creating the awareness that (math-related) intelligence is not a fixed trait but malleable through constant training (Blackwell, 2007).

We analyse the impact of the workshop on the mentioned outcomes by means of a randomized controlled trial with pre-post design. Overall, our results suggest that the workshop fosters students’ psychological resources and slightly increases their intention to enrol in a STEM related major. As hypothesized, this effect is stronger for females. In contrast, there are no visible differences (both in terms of the strength of the effect as well as its statistical significance) between pupils from low and high SES-backgrounds. These results demonstrate that career guidance profits from incorporating socio-psychological methods when tackling gender gaps.

We employ a Randomized Controlled Trial (RCT) to estimate the causal impact of the intervention and to avoid biased treatment effect estimates due to endogenous selection. To this end, we actively recruit participants in two big German cities to participate in our study. The target group are pupils in higher secondary education who will obtain their university entrance diploma within the coming one and a half year. After registration for the study, they completed the first online survey and were randomly assigned to either treatment group or control group. The second wave takes place approximately sixth months after the registration, with the treatment group having attended the counselling workshop between the two waves. The treatment effect estimation relies on the comparison of outcome variables between treatment and control group. To achieve the most efficient estimation, we include socio-economic controls as well as outcomes measures from the first wave into the analysis (Imbens and Rubin, 2015). At least in case of a certain sample size, regression-adjusted techniques are superior to simple differences-in-means comparisons as the control variables help to reduce unexplained variance in the outcome variable. Since we have to cope with two-sided non-compliance as well as a structured dataset (pupils nested in schools/districts), we employ instrumental variable (IV) estimation with clustered standard errors, since non-compliance is likely to be endogenous, therefore creating biased estimates in the same way a non-experimental estimator in case of endogenous treatment selection would. For sake of robustness, we benchmark the results from traditional two-stage least-squares analysis against more flexible semi-parametric estimation techniques (Frölich, 2007), which shows very little differences compared to the baseline specification. The measurement of the psychological outcome variables (academic self-efficacy and growth mind-set) was based on previously used scales that showed good psychometric properties in previous validation studies (De Castella and Byrne, 2015). Since the scales have previously been administered in the American context, we translated them into German and tested them in a short cognitive pre-test. Students’ intended field of study was asked via an open question and was coded according to the International Standard Classification of Occupations (ISCO-08). We then created a dichotomous indicator for STEM related fields.

The paper has been motivated by concerns about a lack of pupils engaging in STEM related fields in general as well as gender and social inequalities in college enrolment and STEM choice in particular. In contrast to previous interventions that mostly focus on tackling information deficits, the treatment under discussion here aims at enhancing students’ psychological resources. The results echo previous studies which have shown that psychological resources can be increased through targeted interventions. Beyond that, we demonstrate that socio-psychological counselling increases the intention to enrol in a STEM related field. While the results for different sub-groups confirm that the intervention may alleviate the gender gap, they do not suggest that it could play a major role in reducing social inequalities with respect to students’ socioeconomic background. In contrast, if high-SES students are more likely to participate in such interventions, they may actually increase educational inequalities. Ultimately, the intervention can be regarded as an effective means to facilitate less anxious-driven and less stereotyped field of study choice for girls, especially given that the costs for this counselling intervention are rather limited. It could therefore effectively complement previous approaches. However, it should be noted that, as typical for RCTs, the external validity could be limited. Due to the way participants have been recruited, the composition of the treatment group may differ from large-scale classroom interventions. Even though the recruitment has taken place through different channels including direct contacts with schools, participation in the study has clearly been on a voluntary basis. It should therefore be noted that an up-scaling of the intervention would go along with a change of the target group. To increase the external validity of our findings, the next step should be to administer a similar treatment in a larger context, e.g. a RCT where randomization takes place at the school level.

Barone, C.; Schizzerotto, A.; Abbiati, G.; Argentin, G. (2017). Information Barriers, Social Inequality, and Plans for Higher Education: Evidence from a Field Experiment. European Sociological Review, 112, 84-96. Barone, C.; Schizzerotto, A.; Assirelli, G.; Abbiati, G. (2018). Nudging gender desegregation: a field experiment on the causal effect of information barriers on gender inequalities in higher education. European Societies, 33, 1–22. Blackwell, L., Trzesniewski, K., Dweck, C. S. (2007). Implicit theories of intelligence predict achievement across an adolescent transition: A longitudinal study and an intervention. Child Development, 78, 246–263. Bettinger, E. P.; Long, B. T.; Oreopoulos, P.; Sanbonmatsu, L. (2012). The Role of Application Assistance and Information in College Decisions: Results from the H&R Block Fafsa Experiment. Quarterly Journal of Economics, 127, 1205–1242. De Castella, K.; Byrne, D. (2015). My intelligence may be more malleable than yours: the revised implicit theories of intelligence (self-theory) scale is a better predictor of achievement, motivation, and student disengagement. European Journal of Psychology of Education, 30, 245-267. Correll, S. J. (2001). Gender and the Career Choice Process: The Role of Biased Self‐Assessments. American Journal of Sociology, 106, 1691–1730. Dweck; C. S. (2000). The development of ability conceptions. In: Wigfield, A., Eccles, J.S., editors. Development of achievement motivation. A volume in the educational psychology series. San Diego, CA: Academic Press; 57–88. Ehlert, M.; Finger, C.; Rusconi, A.; Solga, H. (2017): Applying to college: Do information deficits lower the likelihood of college-eligible students from less-privileged families to pursue their college intentions?: Evidence from a field experiment. Social Science Research, 67, 193–212. Frölich, M. (2007): Nonparametric IV estimation of local average treatment effects with covariates. Journal of Econometrics, 139, 35-75. Imbens, G.W.; Rubin, D.B. (2015). Causal Inference for Statistics, Social and Biomedical Sciences. Cambridge: Cambridge University Press. Leslie S.J.; Cimpian A, Meyer M, Freeland E. (2015). Expectations of brilliance underlie gender distributions across academic disciplines. Science, 347, 262-265. Perez-Felkner, L.; Nix, S.; Thomas, K. (2017). Gendered Pathways: How Mathematics Ability Beliefs Shape Secondary and Postsecondary Course and Degree Field Choices. Frontiers in Psychology 8, 386. Reuben, E.; Wiswall, M.; Zafar, B. (2017). Preferences and Biases in Educational Choices and Labour Market Expectations: Shrinking the Black Box of Gender. In: Economic Journal, 127, 2153-2186. Zafar, B. (2013): College Major Choice and the Gender Gap. Journal of Human Resources 48, 545-595.