The Empowerment Leadership and Mentorship (ELM) program: enhancing academic performance and retention through social-emotional learning and peer mentorship

Higher education institutions are implementing student success strategies with varying results. Historically, support structures have been academic and bureaucratic, focusing on resources like tutoring and office hours. However, without intentional mentorship and care, these structures alone don’t guarantee student success. Institutions can prioritize external metrics like aggregate grades over students’ social-emotional needs, fostering an impersonal environment. The environment and expectations that an institution has on people form the basis of the cues they follow, including when indifferent or negative (Borden & Holthaus, 2018; Gao, Zino, & Ye, 2024; Iacopini, Petri, Baronchelli, & Barrat, 2022; Li, Zheng, Harris, Liu, & Kirkman, 2016).

Positive interactions with academic and student success advisors are crucial (Hawthorne, Zhang, & Cooper, 2022), but essential student services like advising, career counseling, and financial aid are often siloed, leading to poor coordination, workloads, and potentially negative outcomes for students. Meanwhile, students support each other through existing social structures (Borden & Holthaus, 2018), which also has significance due to socioeconomic differences between faculty and students (Collier, 2017; Morgan et al., 2022). Faculty members often come from backgrounds with more familial preparation and support for higher education. This contrasts with the increasing number of students entering universities from less advantaged backgrounds (Morgan et al., 2022).

Peer-mentor-based success strategies show potential (Tucker, Sharp, Qingmin, Scinta, & Thanki, 2020), but their success varies based on organization and execution (Collier, 2017; Venegas-Muggli, Barrientos, & Álvarez, 2021). This peer-to-peer support aligns with social learning theory, which emphasizes the role of observation and interaction in learning (Borden & Holthaus, 2018). Since students will support each other regardless of institutional structures, we have taken the approach that by offering training freely to students in fundamental methods of peer support that benefit student success, this pervasive ad hoc support will improve and start to integrate with the hierarchical structures of the institution (Gao et al., 2024; Iacopini et al., 2022). Successful interactions within such an Empowerment, Leadership, and Mentorship (ELM) program, are characterized by students actively seeking guidance from mentors, mentors effectively facilitating study groups, and collaborative problem-solving. These behaviors not only support stronger academic performance, they are encouraged in the context of the development of emotional intelligence as a basis for building resilient long term learning and interpersonal skills.

Prior research suggests that EI is a multifaceted construct with intercorrelated mental abilities that can develop (and be developed) through targeted training (Mayer, Caruso, & Salovey, 1999). Specifically, approaching EI as a learnable skill could be of tremendous value in educational settings, where social and emotional learning has been found to contribute to improved academic outcomes and personal growth (Greenberg et al., 2003). To this end, the ELM program at Illinois Tech utilized a peer-led implementation strategy to train students in these multiple components of EI as they are applied in the curricular and co-curricular experience of the students, thereby improving interpersonal skills, decision-making, academic performance, and overall collegiate success.

Such a peer learning model is accessible and powerful when implemented strategically. This involves recognizing the connection between social-emotional support and academic goals, and scaling the model while maintaining professional oversight (Gao et al., 2024; Goleman, Boyatzis, & McKee, 2013; Iacopini et al., 2022). Consequently, we developed a multifaceted approach involving coaching and goal setting by staff but with a focus on a semi-autonomous, peer-led structure where students manage their own schedules and agendas. Nevertheless, foundational practices are necessary to establish, maintain, manage, and grow a student success ecosystem.

The Student Success Ecosystem, as implemented in the ELM program at Illinois Tech, is a network of interconnected support structures designed to foster holistic student development and cultural curation first established in the Fall semester 2019 through the office of Academic Affairs in partnership with departments across the institution. It encompasses peer mentorship, academic support, social-emotional learning, and leadership development, creating a cohesive and supportive environment. This peer-mentor-led student support model is a scalable train-the-trainer ecosystem, with practices overlapping common training, communication pathways, overlapping personnel, shared vision, empathetic communication, and receptiveness to suggestions, creating a well-rounded support practice.

ELM utilizes a train-the-trainer model to effectively scale peer mentorship led student success practices (Bastable, 2021; Broad & Newstrom, 1992). This model empowers a core team to develop a large network of student mentors across three implementation tracks, organically spreading peer-to-peer assistance. Students learn to “reframe” core skills, transferring them to diverse contexts and enhancing resilience, flexibility, and stress reduction (Clark, 2013; Ericsson, Hoffman, & Kozbelt, 2018; Nguyen, 2021). ELM training focuses on:

1. Empowerment, Leadership, and Mentorship: Social-emotional intelligence.

2. Metacognition for Learning: Academic and professional mastery.

3. Metacognition for Leadership: Teamwork and community engagement.

This study assesses ELM’s impact on student success at Illinois Institute of Technology.

This study uses the semester-based Grade Point Average (GPA) and persistence measures as quantifiable metrics to assess program outcomes. GPA data was collected from student records for specific cohorts: classes with and without student Supplementary Instruction (SI), students in General Learning Strategies (GLS) courses and/or on probation, and students in scholarship organizations. Data was collected for each semester, excluding Fall 2020 due to pandemic-related disruptions to letter grades. Descriptive statistics were used to compare GPA across these groups. Institutional Review Board (IRB) approval was obtained, and student data was de-identified and aggregated. The use of GPA and retention rates as indicators of student success and program effectiveness is supported by research in similar educational contexts (e.g. Richardson, Abraham, & Bond, 2012; Seidman, 2018; Wang, Dai, & Mathis, 2022).

Trained and supervised mentors support students in challenging or bottleneck courses, fostering independence and study groups (approx. 1,000 students/semester).

Develops transferable skills for academic success and good standing via coaching and peer mentoring.

ELM principles spread organically, with student leaders receiving training and mentorship, building a supportive network. The ELM team provides nominated and self-nominated training to student leaders through workshops and one-on-one mentoring, creating a strong network of supportive leaders, followers, and allies.

The Institutional Review Board (IRB) at the Illinois Institute of Technology reviewed the study proposal and deemed the research and the data presented in this study as exempt from IRB approval. No personal identifying information is disclosed in this manuscript.

Restructuring the Academic Resource Center (ARC) to focus on active hours rather than paid waiting time resulted in a near doubling of supported scholars, sections, and classes while maintaining a flat budget. The data-driven approach deployed resources to classes with higher failure rates or low GPAs. This led to support for most 100 and 400-level courses, with experienced scholar mentors prioritized for “bottleneck” STEM courses like math, physics, chemistry, and computer science that have been identified as classes that present a challenge for a significant proportion of the students taking them.

The mean and standard deviation difference between class GPA with and without SI support was found to be 0.03 across all supported sections which could imply no difference, on aggregate. However, this aggregate view masks significant improvements in key areas. The lowest class GPA with SI support was 0.09 above that of non-SI-supported classes, indicating a baseline improvement for even the most challenging courses. Furthermore, the GPA difference improves radically when SI’s are deployed in specifically identified key classes, such as in the core curriculum STEM, the first classes for majors programs, and advanced (300–400 classes) (Figure 1a). Specifically, Figure 1a demonstrates a half-letter grade improvement in these key courses, suggesting a profound impact on student academic performance and potential career trajectories. This improvement is particularly significant in bottleneck courses, which are often critical for student progression in STEM fields.

We saw significant GPA improvement in classes with SIs in direct collaboration with instructors (Figure 1b). Figure 1b shows a positive shift from failing grades to passing grades, particularly in classes where math support addressed fundamental learning gaps. This suggests that targeted interventions, focusing on addressing underlying academic weaknesses, can effectively improve student outcomes. This is not simply a matter of boosting grades, but of building a solid foundation for future academic success.

In addition to mentor support, this track builds skills in personal leadership, learning how to learn and manage a healthy life-work balance. We saw an exponential increase in the students taking this class. < 10 students prior to 2019 to more than 120 in the Spring 2023 semester, with ∼1 letter grade improvement before/after the class (Figure 2a). This growth in enrollment indicates a strong student demand for the skills and support offered by the GLS course. The student grades continued to improve in the semesters following the class, consistent with a sustained integration of the skills developed. This followed a 0.78-term GPA improvement in the first semester, with some students having much more significant grade improvements.

The percentage composition of students on probation in the class varied from approximately 40 to 80%. With little difference between the rate of GPA improvement between those who started the class in “good standing” and those who started on probation. The return to good standing rate was ∼73% in the first semester and >85% in 2 or more semesters following (Figure 2b). These results, as shown in Figure 2b, indicate a high success rate in helping students return to good academic standing, suggesting the effectiveness of the GLS course in improving student academic performance and retention. The consistency of improvement, regardless of initial academic standing, highlights the broad applicability of the course’s strategies.

Quantifying Track 3’s relative success for the general student population is challenging. We elected to analyze three scholarship organizations as models of the student population. Two actively participated in ELM under mentorship (groups A and C), while the third had similar staff support but lacked a peer mentor basis (group B). This comparison allows us to isolate the impact of the peer mentorship approach within the ELM framework.

Figure 3a shows a clear divergence in GPA trends. Group A, which received closely supervised ELM mentorship, demonstrated a significant upward trajectory, exceeding the school average. This suggests that the approach taken with the ELM program had a substantial positive impact on their academic performance. Conversely, Group B, despite having staff support, saw a decline in GPA, indicating that staff support alone may not be sufficient for maintaining or improving academic outcomes. Group C, which had ELM mentorship in that the cohort leadership was trained in the program, maintained a stable GPA, suggesting that the program supports consistent academic performance in both the staff-supervised train-the-train modality and in the more autonomous student led mode. Figure 3b further illustrates these trends, with Groups A and C showing increases in the percentage of students achieving top grades (A and B), while Group B experienced a slight decrease. This reinforces the idea that peer mentorship plays a key role in enhancing academic achievement.

Participating groups A and C showed higher GPA improvement over time, while group B declined (Figure 3). Fewer students across all groups had lower GPAs, likely due to scholarship requirements. Notably, group A, with many Pell-eligible scholars, increased their professional placement rate from 60 to 100%, matching the high rates of other scholarship groups (∼90–100%). This remarkable increase in professional placement for Group A suggests that the ELM program not only enhances academic performance but also prepares students for successful transitions to professional careers.

While GPA is an imperfect and indirect indicator of learning, it can signal improvements in EI and metacognitive skills that influence academic performance (Honicke & Broadbent, 2016; Richardson et al., 2012). Similar to using temperature as a general indicator of health, GPA may provide a broad assessment, with limitations in precision. GPA reflects more than content knowledge; it’s a complex but aggregate endpoint measurement of the interplay of time management, emotional regulation, and learning strategies. ELM enhances these through EI and metacognition, positively impacting GPA. Therefore, in the context of carefully compared student cohorts, GPA serves as a proxy for program efficacy. Group A’s success, particularly for first-generation, low-income students, demonstrates ELM’s ability to address equity gaps and foster a supportive environment. The fact that Group A (first-generation, low-income students) has reached the school’s average GPA is particularly encouraging as it demonstrates the program’s effectiveness in addressing possible gaps in access and opportunity. This achievement suggests that ELM is not only about improving grades but about fostering a supportive environment where all students can thrive meritoriously. Similarly, Group C continued to improve its upper performance alongside this trend. These are significant indicators of the approach being impactful. For support that is not under explicit full-time staff control, but rather is student-led.

The results presented demonstrate that the ELM program, across its diverse tracks, significantly enhances student academic performance and retention. The strategic restructuring of the ARC, the implementation of the GLS course focused on social-emotional intelligence and metacognition, and the targeted mentorship provided to scholarship groups collectively contributed to notable improvements in student outcomes. These findings underscore the critical importance of integrating holistic support structures, encompassing social-emotional learning, metacognition, and peer mentorship, into the fabric of higher education institutions.

Overall, the ELM program empowers students to translate good intentions into action, partly due to its focus on emotional intelligence development, emphasizing the personal benefits of mutual support. This is evident in the successful implementation of peer-led initiatives, which foster a sense of shared responsibility and collaboration.

The study acknowledges several limitations. Firstly, the reliance on GPA as an indirect measure of learning necessitates cautious interpretation. While GPA serves as a valuable proxy, it does not fully capture the nuanced development of EI and metacognitive skills. Future research could incorporate qualitative assessments, such as student interviews and reflective journals, to gain a deeper understanding of these skills. Secondly, quantifying the impact of mentorship on the general student population presents inherent challenges. Future studies could employ longitudinal designs to track the long-term effects of mentorship and explore the specific mechanisms through which EI and metacognitive skills influence academic performance. Additionally, network analysis could be used to see how the ELM program changes the student social networks.

These findings align with broader literature on student success, which emphasizes the importance of social-emotional factors and peer support in fostering academic achievement (Kuh, Kinzie, & Buckley, 2006; Tinto, 1993). The ELM program’s success in supporting diverse student populations, particularly first-generation and low-income students, highlights its potential to address opportunity and access gaps in higher education. Future research could explore the scalability of this model to other institutions and investigate its applicability across different disciplines and student demographics. Finally, the peer mentor-based approach to student success is likely to be applicable in various other aspects, including professional practice and voluntary activities for people of all ages and occupations. This suggests that the principles of ELM can extend beyond the academic setting, contributing to personal and professional development in diverse contexts.

By considering these suggestions, leadership educators can adopt and adapt the ELM model.

Start Small and Scale Gradually: Begin by piloting the ELM approach with a small group of students and mentors. This allows for refining the program to/for your community and addressing potential challenges before expanding to a larger population. The initial mentors significantly affect the outcome at scale, making this stage particularly critical.

Invest in Mentor Training: Provide comprehensive training to peer mentors, encouraging them to take ownership of their roles as partners, not just participants. Emphasize active communication skills, empathy, community-building, self-care, and leadership involving accountability and feedback.

Foster a Supportive Environment: Create a culture that encourages open communication, trust, and mutual respect among students, student mentors, and mentors who are staff, faculty, or administrators. Be honest and sincere, and manage expectations in the same spirit.

Data-Driven Decision Making: Continuously collect and analyze data to track student progress, assess program effectiveness, and inform adjustments to the ELM model. It’s tempting to overanalyze data, but remember that the most important choices are often simple after some reflection. Focus on understanding what you are trying to measure. For us, it took time to determine that a simple semester GPA provided a non-biased quantitative measure of impact. It’s an elegantly simple metric for a university – it is the single most common measure of a student’s integration with their purpose at the institution. This might not be the case for your implementation and our suggestion is to find the quantifiable measure/s that are most appropriate with minimal prior assumption.

Secure Institutional Buy-in: Gain support from faculty, staff, and administrators to ensure the program has the necessary resources and integration into the existing student support structure. However, don’t try to launch with a grand scheme; remember the first point of starting small and building sustainably. This might mean choosing an initial area to start the program that can grow naturally and include others as it scales, such as student organization platforms, clubs, scholarships, and so forth. Include faculty, staff, and administrators as mentors organically, based on their genuine inclination and commitment to be engaged rather than feeling obligated.

Funding: This study was funded by the Robert A. Pritzker Endowment, supporting the ELM program at Illinois Tech.