One of the most critical sources of socio-economic inequality is represented by educational attainment. It is likely to foster long-term effects on a wide set of outcomes including income, wealth, and health. Underrepresented minorities by gender and race are more likely to have lower educational performance than non-minority students. STEM fields (Science, technology, engineering, and mathematics) offer very promising careers, better hiring prospects, and substantial salary premiums. According to OECD (2019) data, in member countries, fewer than 1 in 3 engineering graduates and fewer than 1 in 5 computer science graduates are females, which is frequently due to stereotypes and expectations, rather than performance differences in math and science. University instructors and lecturers are likely to provide decisive role models and this is especially true in the case of female or ethnic minority students in STEM fields. The first aim of this research project is to explore interactions of minority students with the same minority lecturers and instructors in a vast array of STEM courses at the Sapienza University of Rome. The project will assess to what extent exposure to the same minority instructor impacts students¿ performance in STEM disciplines. In order to circumvent the issue of possible self-selection of students to courses and instructors, the project is set to exploit an uncommon assignment method of students to instructors based on the mechanism of ¿canalizzazione¿, where students are allocated to different instructors based on the initial of their surname, constituting a random assignment mechanism. The second aim of the project is to set up a randomized field experiment to test if de-biasing outreach letters sent via e-mail to randomly selected students representing the aforementioned minorities containing the information about the carrier prospects have an impact on their academic outcomes.
The innovative aspects of this research project are as follows:
1) This project would represent the first study to exploit individual data from the largest European university in order to provide evidence on the importance of role models for minority students in STEM fields. The existing studies are based on limited sample sizes collected in particular settings, while the unprecedented dataset dimension, in this case, will enable us to conduct a more thorough analysis, allowing us to account for important sources of heterogeneity that might drive the results.
2) The study will contribute to the so-far scant literature on ethnic minorities in STEM fields in tertiary education. To our knowledge, the most recent study is that of Oliver et al. (2021), who provide evidence on whether underrepresented minority students experience achievement gains from being taught by underrepresented minority teaching assistants in STEM. The second study of the effect of minority instructors on STEM disciplines is that of Price (2010), who shows that first-term black students enrolled at public US universities are more likely to persist in STEM when they are exposed to same-race instructors. A partially close paper to the two mentioned is also that of Lusher et al. (2018), who find that economics students perform better in classes taken with teaching assistants who are of a similar race.
3) This study would be one of the first ones to use the RCT technique in STEM fields in tertiary education. For earlier educational levels, a number of psychological studies examine the impact of role models in STEM fields, especially from a gender perspective, building on the evidence that STEM disciplines are labeled unfeminine. For example, Betz and Sekaquaptewa (2012) show that counter stereotypic-yet-feminine success may actually be demotivating, particularly to young girls, and find that feminine STEM role models reduced middle school girls' current math interest, self-rated ability, and success expectations relative to gender-neutral STEM role models, calling for a better understanding of feminine STEM figures aimed at motivating young girls. On the other hand, Cheryan et al. (2011) show that exposure to a non-stereotypical STEM role model ("e.g. someone who dresses fashionably, enjoys sports and hanging out with friends, and watches "normal" TV shows, such as The Office") through a "getting to know each other" task in the lab increases female subjects' beliefs of succeeding in the field and that this is true regardless of the gender of that role model. In fact, according to the authors, women's perception of characteristics associated with STEM majors such as isolation, excessive focus on computers, and unsociability is not compatible with how they perceive their societal role, frequently involving helping others and preserving physical appearance. From that viewpoint, their findings suggest that non-stereotypical role models, whether man or woman, are likely to impact females' preferences and beliefs.