Female peer mentors early in college have lasting positive impacts on female engineering students that persist beyond graduation

The underrepresentation of women in STEM fields, particularly engineering, is a significant concern. The demand for skilled STEM workers is high, and the gender gap exacerbates existing income inequality. While interventions focusing on cognitive reappraisal, such as self-affirmation and growth mindset, have shown some success in reducing achievement gaps, they have limitations. These interventions place the onus on individuals to adapt rather than on institutions to change. Furthermore, many focus on grades and test scores, neglecting subjective experiences like belonging and confidence, which are crucial for persistence. Another approach involves increasing the visibility of successful role models. Research suggests that interacting with successful same-gender role models can positively impact performance and subjective experiences. However, most studies assess outcomes at a single time-point or over short periods, failing to capture long-term effects and the impact of authentic mentoring relationships. This study builds on prior research by examining the long-term effects (through graduation and one year post-graduation) of a year-long peer mentoring intervention for female engineering students. The focus on female mentors in engineering is particularly relevant due to the low representation of women in these fields. This research uses a randomized controlled trial to determine the causal effect of peer mentoring, addressing limitations of previous studies where participation was self-selected or mentors were chosen by mentees. This study specifically predicts that authentic mentoring relationships with same-gender role models will have substantial long-term benefits, bolstering confidence, motivation, and persistence in STEM.

The study reviewed existing literature on interventions to address underrepresentation in STEM. It highlighted the limitations of mental reappraisal interventions, which often focus on individual adaptation rather than institutional change, and primarily measure objective outcomes like grades. The importance of role models and their impact on performance and subjective experiences was discussed. However, the literature lacked longitudinal studies on authentic mentoring relationships and the potential long-term impact of same-gender mentorship in real-world settings. This study aimed to fill this gap, examining the long-term effects of a year-long peer mentoring intervention on female engineering students, comparing the effects of female and male mentors against a control group. The existing research suggested that mentoring relationships between students and mentors sharing marginalized identities would likely be more impactful than brief exposure to such individuals.

A longitudinal randomized controlled field experiment was conducted with 150 first-year and transfer female undergraduate engineering students at a large public university. Participants were randomly assigned to one of three conditions: female peer mentor (n=52), male peer mentor (n=51), or no mentor (n=47). Mentors were junior or senior engineering students who received training before being matched with mentees. The study was described to both mentors and mentees as focusing on barriers and opportunities for engineering students, maintaining participant blindness to the study's focus on gender. Mentor-mentee dyads met a median of four times during the mentees' first year. Mentees completed surveys at multiple time points: baseline (before mentor assignment), mid-year, end-of-year, and annually through graduation and one year post-graduation. Data included subjective experiences (motivation, confidence, aspirations, emotional well-being), objective outcomes (internship participation, choice of major), and college transcripts (to determine major and GPA). Multilevel models in Mplus 8 were used to analyze the longitudinal data, handling missing data using full information maximum likelihood. Chi-square tests analyzed dichotomous variables. Mediation analyses examined whether changes in confidence, motivation, or emotional well-being mediated the effect of female mentorship on academic choices.

Compared to the male mentor and no-mentor conditions, assignment to a female mentor was associated with several significant positive outcomes. Participants with female mentors showed: * **No decline in confidence:** Unlike the other two groups, their confidence in engineering abilities remained stable throughout the study period. Participants in the no-mentor and male-mentor conditions experienced a significant decline in confidence. * **Maintenance of motivation:** Motivation in engineering courses remained stable for those with female mentors, while the other two groups experienced a significant decrease. * **Higher internship participation:** A significantly higher percentage of participants with female mentors (82%) participated in engineering internships compared to those with male mentors (65%) or no mentor (61%). * **Greater STEM degree completion:** Although engineering degree completion didn't differ significantly, a significantly higher percentage of those with female mentors (92%) graduated with a STEM degree compared to those with male mentors (78%) or no mentor (81%). * **Preservation of postgraduate aspirations:** Intentions to pursue postgraduate degrees in engineering showed less decline for those with female mentors compared to the other groups. * **Stable emotional well-being:** Unlike the other groups, who experienced a decline in emotional well-being, those with female mentors maintained stable levels. Mediation analyses revealed that changes in confidence in engineering skills mediated the effects of female peer mentors on academic choices (internships, STEM major, and postgraduate aspirations).

The findings demonstrate that a brief, low-cost peer mentoring intervention with a same-gender mentor can have a significant and lasting impact on female engineering students' success and well-being. The results support the stereotype inoculation model, showing that female peer mentors can act as 'social vaccines,' protecting against negative stereotypes and boosting confidence and motivation. The fact that these benefits persisted beyond the intervention period and even after graduation highlights the long-term impact of such interventions. The mediating role of confidence suggests that fostering confidence in engineering abilities is crucial for promoting persistence and success in the field. The lack of significant differences in GPA across conditions suggests that the positive effects were not solely due to improved academic performance, but rather influenced by improved subjective experiences and access to resources and support.

This study provides strong evidence for the effectiveness of female peer mentoring in improving the long-term academic and psychological outcomes of female engineering students. The findings highlight the importance of relational interventions, particularly those involving same-gender mentors, in creating more supportive and inclusive environments within STEM. Future research could replicate these findings in other STEM disciplines and explore the effectiveness of similar interventions at different career stages, including graduate school and early career development. It's also important to expand research to other underrepresented groups in STEM.

The study was conducted at a single university, limiting the generalizability of the findings. The sample size, while substantial, might not fully capture the diversity of experiences among female engineering students. Although the researchers attempted to control for pre-existing individual differences by assigning mentors randomly, some unmeasured individual characteristics might still have influenced the results. The self-reported nature of many measures introduces the possibility of response bias. Future work should explore these limitations to confirm results.