The power of extracurricular activities in shaping future leaders Open Access

Leadership is increasingly understood not as an innate trait but as a learned, developmental capacity that emerges through experience, reflection, participation in curricular and co-curricular programs, and engagement in intentional learning environments (Traini, Qgwiji, Velez, & Ross, 2025; Walzer, 2024). Participation in extracurricular environments is widely associated with student learning and leadership development (Astin, 1984; Kuh, 1995, 2009), yet much of this literature remains cross-sectional and concentrated in Western contexts.

The present study addresses these gaps by examining leadership development across the first academic year through students' participation in extracurricular activities (ECAs) at a university in a post-Soviet context. Drawing explicitly on Astin's (1984, 2014) Input–Environment–Outcome (I–E–O) framework, leadership is conceptualized as a developmental behavioral capacity shaped by the interaction between student inputs and educational environments. Specifically, leadership refers to observable tendencies toward initiative, responsibility, social influence, self-regulation, and decision-making in group contexts. Leadership development is therefore understood as change over time in these capacities, reflecting students' evolving ability to act, influence, and make decisions within educational and group settings. In this study, leadership is thus treated not as a formal position or fixed personality trait, but as a learnable educational outcome that may develop through structured experience.

Although prior research links extracurricular participation to communication skills, teamwork, initiative, and leadership capacity, it remains unclear which types of extracurricular environments most effectively support leadership development, particularly in transition-economy contexts. In addition, leadership education research has relied heavily on instruments developed in Western settings, potentially limiting their contextual validity. To address this limitation, the study employs the Zharikov–Krushelnitsky leadership diagnostic, a culturally grounded instrument validated in post-Soviet populations, enabling a contextually appropriate examination of leadership learning in an underrepresented educational setting.

This study contributes to leadership education by examining how leadership development in the first year of university is associated with structured extracurricular environments and evolving student autonomy. Because ECA engagement is measured as participation breadth and exposure rather than intensity, quality, or leadership role, the findings are interpreted as evidence of associations with extracurricular environments rather than definitive evidence of developmental effects.

Early leadership theories conceptualized leadership as an innate and biologically inherited trait (Spector, 2016). Subsequent empirical research, however, failed to identify a consistent set of traits that reliably predict leadership effectiveness across contexts (Hoffman, Woehr, Maldagen-Youngjohn, & Lyons, 2011). Contemporary scholarship conceptualizes leadership as learned, developmental, and socially embedded rather than innate or position-based (McCauley & Palus, 2021; Owen, 2012). Consistent with the definition introduced above, leadership is treated here as a behavioral, developmental capacity shaped through interaction between student inputs and educational environments. This definition aligns with Wilson's (2023) view of leadership as purposive social influence and a developmental outcome rather than a fixed trait. Within Astin's I–E–O framework, leadership is therefore treated as a contextually embedded outcome shaped by the interaction between student characteristics and developmental environments.

Astin's (1984, 2014) Input–Environment–Outcome (I–E–O) framework provides a foundation for examining leadership development in higher education (Figure 1). The model distinguishes among student inputs (demographic and academic characteristics), environments (educational experiences), and outcomes (developmental changes such as leadership).

Astin (1984, 2014) defined student involvement as the physical and psychological energy students invest in academic and social activities. Within this perspective, ECAs serve as important environments for leadership learning, particularly when formal leadership education is limited. Research grounded in the I–E–O framework links ECA participation to academic engagement, psychosocial development, employability skills, and leadership outcomes (Ribeiro, Malafaia, Neves, & Menezes, 2024).

ECAs provide structured opportunities for collaboration, decision-making, and leadership practice (Bodolica, Spraggon, & Badi, 2021) and are associated with competencies such as confidence, teamwork, and initiative (Kim, 2022). Astin's framework further suggests that learning-related outcomes may depend on the quality, structure, and developmental fit of involvement rather than participation alone. In this study, extracurricular engagement is operationalized as the breadth of participation across activity types, capturing exposure to diverse extracurricular environments within Astin's I–E–O framework without directly observing the quality or intensity of involvement.

Extracurricular engagement was categorized into domains representing distinct institutional environments for leadership development. Prior research conceptualizes involvement as participation across different organizational contexts rather than a single activity type (Astin, 1984; Kuh, 1995, 2009). Consistent with this perspective, Rosch, May, Wilson, and Spencer (2023) identify several common forms of co-curricular engagement, summarized in Table 1.

The ECAs available at Westminster International University in Tashkent (WIUT) broadly align with the domains identified by Rosch et al. (2023) and reflect the university's extracurricular environment. These activities are listed in Table 2.

The activities in Table 2 were grouped into categories representing different developmental environments. Intellectual clubs and academic games emphasize analytical thinking and discussion. Language clubs promote intercultural communication and global awareness. Artistic and creative clubs (e.g., music, art, photography) support creative expression and collaboration. Sports were divided into team and individual activities to distinguish leadership experiences based on collective coordination versus personal performance and self-regulation. This classification allows examination of how different types of involvement contribute to leadership development within the university environment.

Demographic and academic characteristics influence whether students participate in ECAs and the types of activities they enter. Gender has been shown to predict ECA participation patterns, with males more likely to engage in competitive and athletic activities and females more likely to join academic and arts-based clubs (Johnson & Routon, 2024; Kim, 2022). Academic preparation further structures engagement. Students with stronger academic performance and higher language proficiency tend to exhibit greater institutional integration and confidence, facilitating participation in cognitively demanding and communication-intensive environments (Kuh, 1995, 2009). English proficiency is particularly salient in internationally oriented extracurricular contexts.

Age and academic maturity also shape engagement trajectories. Younger students often explore a broader range of ECAs, whereas older or academically constrained students engage more selectively as responsibilities increase (Reason, Terenzini, & Domingo, 2010). Accordingly, we hypothesize:

Astin's framework also allows for direct Input → Outcome relationships, whereby student characteristics influence developmental outcomes independently of environmental engagement (Astin, 1984, 2014). Prior research links gender to leadership emergence and leadership self-efficacy during early university stages (Eagly & Carli, 2007; Kim, 2022). Age and maturity relate to self-regulation and social awareness, competencies associated with leadership development (Komives, Longerbeam, Owen, Mainella, & Osteen, 2005).

Academic preparation also plays a role. Language proficiency facilitates communication and influence, while stronger academic performance supports problem-solving, confidence, and persistence (Kuh, 2009; Reason et al., 2010). Leadership development in higher education emerges from students' interactions with developmental environments that offer opportunities for involvement, responsibility, and reflection (Astin, 1984). From a leadership education perspective, these inputs shape students' readiness for leadership learning.

Participation in ECAs provides contexts for responsibility, collaboration, decision-making, and social influence (Bodolica et al., 2021). Different types of ECAs may support leadership learning through distinct mechanisms. Intellectual games emphasize analytical thinking and strategic decision-making (Sousa & Rocha, 2019); sports foster discipline, resilience, and teamwork (Taras, 2005); music and arts activities promote emotional regulation and sustained practice, while language clubs strengthen communication confidence (Yusof & Abugohar, 2017). Student organizations offer important opportunities for leadership development, identity formation, and collaborative learning (Komives et al., 2005). Participation in such activities is linked to higher motivation, interpersonal competence, academic engagement, and leadership development (Foreman & Retallick, 2013; Hancock, Dyk, & Jones, 2012; Kim, 2022; Seow & Pan, 2014). Accordingly, we hypothesize:

A central proposition of Astin's I–E–O framework is that educational environments mediate the relationship between student inputs and developmental outcomes. Academic preparation may be related to leadership development indirectly by shaping access to and persistence in leadership-relevant environments.

Students with stronger academic backgrounds are more likely to engage in structured ECAs that provide opportunities for leadership practice and social influence (Reason et al., 2010). These environments, in turn, foster leadership competencies such as initiative, communication, and self-efficacy (Komives et al., 2005). Prior research suggests that extracurricular involvement partially mediates the relationship between academic ability and leadership development (Kuh, 1995, 2009). Therefore, we hypothesize:

Astin's framework also allows for conditional effects, in which the impact of environments varies across student subgroups. Gender norms and living arrangements may shape how students experience and benefit from extracurricular environments. Gender differences in participation and leadership development are well documented (Kim, 2022). Living independently may further enhance autonomy and responsibility, both of which are closely aligned with leadership learning, because they require self-regulation and decision-making beyond institutional supervision (Hamzah, Baharuddin, Fauzi, & Aripin, 2019). Accordingly, we hypothesize:

Astin's I–E–O framework supports a longitudinal view, proposing that early engagement with educational environments can generate cumulative developmental effects. Leadership competencies emerge through sustained practice and reinforcement rather than single experiences (Komives et al., 2005; McCauley & Palus, 2021).

Early ECA participation may therefore be associated with the emergence of leadership development trajectories that influence outcomes later in the academic year. Longitudinal research suggests that early engagement predicts later leadership and psychosocial development, even after controlling for baseline characteristics (Reason et al., 2010). Thus, we hypothesize:

This study employed a two-wave longitudinal quantitative design to examine leadership development during the first year of university. Data were collected at the beginning of the academic year (Study 1) and approximately eight months later at the end of the year (Study 2). Consistent with Astin's (1984, 2014) I–E–O framework, the design examines concurrent, mediated, moderated, and longitudinal relationships among student inputs, ECA engagement, and leadership outcomes.

Participants were first-year undergraduate students enrolled at WIUT, an English-medium institution operating in a post-Soviet higher education context. In Study 1, 482 students completed the survey. In Study 2, 185 of the same students participated in the follow-up, yielding a retention rate of 38.4%, which is comparable to rates reported in longitudinal research on first-year university cohorts. Verbal informed consent was obtained prior to data collection.

Leadership was assessed using the Zharikov–Krushelnitsky Leadership Diagnostic Questionnaire, a 50-item dichotomous instrument designed to capture leadership tendencies in applied social contexts (Fetiskin, Kozlov, & Manuilov, 2002). The instrument aligns with this study's behavioral definition of leadership because it measures tendencies related to initiative, responsibility, influence, emotional regulation, and decision-making. These dimensions correspond to the behavioral components of leadership defined in this study, strengthening alignment between measurement and theoretical framing. Rather than assessing formal leadership position or leadership identity, the Z–K diagnostic captures action-oriented dispositions relevant to leadership learning in educational environments. Responses were summed to create a Leadership Index, with higher scores indicating stronger behavioral leadership tendencies.

Internal consistency was acceptable for a broad multidimensional developmental measure in Study 1 (KR-20 = 0.65) and weaker in Study 2 (KR-20 = 0.55), which is acknowledged as a limitation. However, it is consistent with leadership education research that uses broad behavioral indicators rather than unidimensional trait scales (e.g. Komives et al., 2005). The use of the Z–K diagnostic is nevertheless appropriate because it was developed and used in post-Soviet educational and social-psychological contexts (Paramonova, 2020). This supports contextual fit by measuring leadership through locally recognizable behavioral expressions of responsibility, influence, and initiative rather than relying solely on Western-developed leadership scales. This alignment between construct definition, measurement, and cultural context addresses the study's central outcome: leadership as a learned behavioral capacity rather than a formal role or fixed trait.

Extracurricular activity (ECA) engagement was measured through self-reported participation in diverse university-recognized activities (see Table 2). Participants reported both current participation and cumulative years of involvement in each category. For analyses of specific activity environments, participation was coded dichotomously (0 = no participation, 1 = participation). Overall engagement was operationalized as total ECA involvement (years), calculated as the sum of years of participation across activity categories. This operationalization captures breadth of extracurricular exposure across different developmental environments.

Because the survey did not capture hours, quality of participation, or leadership responsibilities, the measure reflects presence and breadth of engagement rather than intensity, depth, or role-based involvement. Involvement-based perspectives emphasize that participation is a necessary condition for learning and leadership development (e.g. Astin, 1984), although outcomes depend on how students are involved. Accordingly, findings reflect exposure to extracurricular environments rather than the quality of participation or leadership roles. Mediation analyses are therefore interpreted as exploratory tests of whether exposure to extracurricular environments is statistically associated with leadership outcomes, rather than as evidence of a fully specified causal mechanism.

Student input variables included gender (0 = female, 1 = male), age (in years), English language proficiency (IELTS score), quantitative ability (mathematics entrance examination score), and academic performance (semester GPA). Living arrangement (0 = living with family, 1 = living independently) was included as a contextual input and examined as a moderator, reflecting its relevance to autonomy and leadership development in higher education settings.

Data were analyzed using SPSS and Hayes' PROCESS macro. Descriptive statistics and correlations were first computed, followed by t tests and regression analyses testing Input → Environment and Input → Outcome relationships (H1–H2) and Environment → Outcome effects (H3), mediation analyses (PROCESS Model 4) examining whether ECA engagement mediated academic inputs and leadership (H4), moderation tests of gender and living arrangement (H5), and longitudinal regressions assessing whether Study 1 ECA participation predicted Study 2 leadership while controlling for baseline leadership (H6).

Statistical significance was evaluated at p < 0.05. Due to listwise deletion, sample sizes varied slightly across PROCESS models (in Appendix Table A5). The longitudinal design allows examination of time-sensitive relationships without implying causality. Leadership was assessed using a culturally grounded, behaviorally oriented instrument appropriate for post-Soviet educational contexts.

Descriptive statistics for all variables are reported in Appendix Table A1. At the beginning of the academic year (Study 1; N = 482), students demonstrated a moderate level of leadership (M = 30.30, SD = 4.96). By the end of the academic year (Study 2; N = 185), leadership scores were modestly lower (M = 28.26, SD = 4.19), a finding that should be interpreted with caution. Leadership scores reflect behavioral tendencies such as initiative, responsibility, and social influence, consistent with the study's conceptualization.

This finding should not be interpreted as a decline in leadership capacity, given attrition and lower reliability at Study 2. It may reflect learning-related context and measurement conditions rather than definitive change in students' leadership. Pearson correlations (Appendix Table A2) showed that environments were differentially related to outcomes. In Study 1, leadership was positively associated with multiple ECA types (rs = 0.16–0.28, ps < 0.01). In Study 2, fewer associations remained significant, with leadership correlated primarily with participation in music and performing arts clubs and independent living.

Hypothesis 1 proposed that student background characteristics would predict ECA engagement (Input → Environment). Independent-samples t tests indicated that in Study 1, male students reported higher total ECA involvement (years) than female students, t(480) = −2.94, p = 0.003, d = 0.29. Students without prior ECA experience also reported marginally higher current involvement, t(480) = 2.09, p = 0.037, d = 0.19. Age, academic inputs, and scholarship status were not significant predictors. Table 3 summarizes all significant group differences.

In Study 2, gender differences approached statistical significance (p = 0.051), while all other inputs were non-significant. Overall, H1 received partial support, with gender emerging as the most consistent predictor of engagement.

Hypothesis 2 examined direct Input → Outcome effects. In Study 1, male students scored significantly higher on leadership than female students, t(480) = −5.38, p < 0.001, d = 0.54. Multiple regression analysis (Table 4) showed that gender (β = 0.21, p < 0.001) and English proficiency (β = 0.10, p = 0.028) significantly predicted leadership, explaining 10.9% of the variance, F(6, 474) = 9.68, p < 0.001 (Appendix Table A3).

By Study 2, these effects were no longer statistically significant. Living independently emerged as the only significant predictor of leadership (β = 0.19, p = 0.014), while gender and academic inputs were no longer significant. Thus, H2 was partially supported, indicating that the importance of student input variables for leadership outcomes shifts across developmental stages.

Hypothesis 3 tested whether specific extracurricular environments predicted leadership outcomes (Environment → Outcome). In Study 1, hierarchical regression showed that adding ECA variables significantly improved model fit beyond inputs (ΔR² = 0.091, F(8, 466) = 6.33, p < 0.001). Participation in intellectual games (β = 0.17, p = 0.005) and individual sports (β = 0.16, p = 0.006) were significantly positively associated with leadership scores (Table 5).

In Study 2, the ECA block did not significantly improve model fit (ΔR² = 0.078, p = 0.062). However, participation in music and performing arts clubs (β = 0.27, p = 0.003) and independent living (β = 0.22, p = 0.004) were positively associated with leadership outcomes, while total ECA involvement (years) was negatively associated (β = −0.34, p = 0.035). H3 was therefore partially supported. Associations between ECAs and leadership differed by activity type and timing, while participation breadth was not a consistent correlate (Appendix Table A4).

Hypothesis 4 tested whether ECAs mediated the relationship between academic inputs and leadership. In Study 1, total ECA involvement was positively associated with leadership (b = 0.10, p < 0.001), and language proficiency showed significant total and direct effects (Table 6). However, in the full mediation model, the indirect effect was not significant (b = 0.0015, SE = 0.0017, 95% CI [−0.002, 0.005]; Appendix Table A5), matching Table 6, where the bootstrap confidence interval for the indirect effect also included zero. The input-to-ECA (a) path was also non-significant in both studies (Appendix Table A5). Only intellectual games showed a small indirect coefficient in Study 1, although this effect was not statistically significant as the confidence interval included zero.

In Study 2, neither the ECA–leadership (b) path nor the indirect effect was significant. Although music and performing arts participation and living arrangement were associated with leadership (Table 6), these effects did not reflect mediation. Thus, H4 received limited support, with weak, context-specific effects that were not sustained over time and should be interpreted cautiously. This suggests that, within this sample, extracurricular engagement did not function as a mechanism linking academic inputs to leadership outcomes.

Hypothesis 5 tested conditional effects. Gender did not moderate the relationship between ECA involvement and leadership (p = 0.833). In contrast, living arrangement significantly moderated this relationship (b = −0.206, p = 0.049), such that ECA involvement was more strongly associated with leadership for students living under supervision than for those living independently (Table 7). Accordingly, H5(b) was supported, while H5(a) was not. Refer to Appendix Table A6 for a more detailed analysis.

Hypothesis 6 proposed that early ECA engagement would predict leadership later in the academic year. Lagged regression analyses showed that total ECA involvement (years) at Study 1 did not predict leadership at Study 2 (b = 0.01, p = 0.763) after controlling for baseline leadership and inputs (Table 8). In contrast, English language proficiency (b = 1.08, p = 0.027) and independent living (b = 2.99, p = 0.009) consistently predicted leadership over time. H6 was therefore not supported.

To contextualize these findings, McNemar's test revealed a significant decline in ECA participation over time, χ²(1, N = 185) = 21.25, p < 0.001 (Table 9). More students discontinued ECAs than began participating, indicating that environmental exposure itself changed over time. This fluctuation in the “Environment” component of Astin's I–E–O model may help explain why early participation did not predict later leadership scores.

Supplementary analyses compared students from Tashkent with those from other regions (see Appendix Table A7). The results showed that regional students were more likely to hold scholarships (χ² = 16.79, p < 0.001), report prior school ECA participation (χ² = 5.50, p = 0.019), and live independently at both time points (χ²s ≥ 6.23, ps ≤ 0.013). A univariate GLM further indicated that independent living predicted leadership in Study 2, F(1, 182) = 4.45, p = 0.036, while multivariate tests confirmed an overall regional advantage, Wilks' λ = 0.945, p = 0.017. These results suggest that regional background may function as a salient student input associated with autonomy-related environments and later leadership scores.

Guided by Astin's (1984, 2014) I–E–O framework, this study examined how student inputs and extracurricular environments are associated with leadership outcomes during the first year of university in a transition-economy context. The findings provide qualified support for the framework and suggest that leadership learning may be developmental and time-sensitive, rather than a simple cumulative function of participation breadth, in line with prior research (e.g. Rosch & Collins, 2019).

Early in the academic year, leadership outcomes were associated with gender, English language proficiency, and participation in structured extracurricular activities, particularly intellectual games and individual sports. These environments are linked to behavioral tendencies related to responsibility, self-direction, and performance under challenge.

As the academic year progressed, the relative importance of inputs and environments associated with leadership shifted. Total extracurricular involvement no longer explained leadership outcomes; instead, autonomy-enhancing conditions, most notably independent living, alongside selective engagement in music and performing arts activities, became more salient. These results suggest leadership outcomes may be related to developmental fit and autonomy rather than breadth of involvement. However, the observed decline in leadership scores should be interpreted cautiously and not treated as direct evidence of reduced leadership capacity. In the post-Soviet higher education context, students often move from highly structured, teacher-directed schooling into university environments that require greater autonomy and self-direction. This context may help explain why structured activities, such as intellectual competitions and individual performance environments, were more salient early, while independent living and selective engagement became more relevant later.

The study contributes to extending Astin's I–E–O framework by suggesting that environmental effects may be developmentally sequenced and context-dependent. Student background factors were connected with early outcomes, while later leadership scores were more closely linked to autonomy-supportive contexts. The longitudinal design and culturally grounded measure support a context-sensitive analysis of leadership development. While suggestive, these patterns should not be interpreted as definitive evidence of developmental change without longer-term longitudinal validation.

Overall, this study advances leadership education by conceptualizing leadership as a learned behavioral capacity shaped by student inputs and educational environments. For leadership educators, the findings suggest the value of extracurricular ecosystems that provide structured challenge early and progressively support autonomy, reflection, and adaptive capacity.

Because the findings are associational and based on breadth of participation, the following implications are offered as cautious design considerations for leadership educators rather than prescriptive causal recommendations. First, communication-intensive environments may support early-stage leadership learning, as leadership scores in Study 1 were positively associated with English language proficiency and participation in intellectually demanding extracurricular activities. Leadership educators may therefore integrate structured activities such as debate and Model United Nations into first-year development pathways. Second, activities emphasizing individual accountability may support behavioral leadership tendencies. Individual sports and later participation in music and performing arts were positively associated with leadership scores, suggesting that sustained effort, personal responsibility, and structured feedback may matter more than participation breadth alone. Third, developmental sequencing may be important. The later salience of independent living suggests that leadership education should gradually scaffold students from structured participation toward more autonomous responsibility, decision-making, and initiative.

This study provides one of the first longitudinal quantitative examinations of leadership development in a post-Soviet higher education context. By integrating Astin's I–E–O framework with an explicit behavioral definition of leadership and a culturally grounded measure (the Zharikov–Krushelnitsky leadership diagnostic), the research extends leadership education scholarship beyond Western settings and highlights the value of time-sensitive developmental analysis.

The study refines Astin's I–E–O framework by suggesting that leadership development depends not only on involvement but also on the type and timing of engagement. Early leadership outcomes were associated with academic readiness and structured challenge, whereas later outcomes were more closely related to autonomy-enhancing environments and selective participation. These findings suggest that leadership development may be nonlinear, shaped by shifting configurations of inputs, environments, and outcomes across developmental stages.

At a broader societal level, the findings highlight how equitable access to high-quality extracurricular environments is associated with leadership development and social mobility, particularly for students from regional or under-resourced backgrounds. By investing in inclusive and structured extracurricular ecosystems, higher education institutions can support the development of socially responsible graduates with leadership capacity who are prepared to contribute to community and national development in transition economies.

Several limitations warrant consideration. First, the single-institution design limits generalizability, and leadership was assessed through self-reported measures rather than observed behavior. Second, attrition between the two waves of data collection and lower internal consistency at the second time point may constrain the strength of some longitudinal findings. Although these limitations are common in longitudinal student development research, future studies may benefit from larger samples and repeated measures to strengthen reliability and reduce attrition.

Extracurricular engagement was measured as participation breadth rather than intensity, quality, or leadership roles. As a result, variation in depth of involvement is not captured. Future research should incorporate more nuanced measures of participation intensity, duration, quality, and leadership responsibility. Multi-institutional and cross-cultural designs, longer longitudinal observation, and mixed-method approaches incorporating peer or instructor assessments and behavioral indicators would further strengthen understanding of how extracurricular environments support leadership development over time.

The supplementary material for this article can be found online