The research team, led by Annabel Wanza Mbithi of Chuka University together with Peter Kimanthi Mbaka and Kenkelvin Kimathi Mbaka, and including Victor Okoth Saoke of the University of Embu, set out to answer a deceptively simple question: what determines how much students say they participate in classroom activities, and do those determinants differ by gender? Participation in this context is not a vague aspiration but a measurable behaviour. It covers the frequency with which students answer questions, contribute to discussions, take part in group work and engage with the practical activities that teachers build into their lessons. Decades of international research have linked such engagement to achievement, retention and eventual transitions into further education and employment, making the classroom voice of individual students a surprisingly consequential variable.

To capture it, the researchers employed a cross-sectional descriptive survey design, a standard tool for taking a snapshot of a population at a single point in time. Working within public day secondary schools, schools that students attend during the day while living at home, they sampled 320 Form Three students, typically the penultimate year of Kenyan secondary education. The sampling strategy was both multistage and proportionate, meaning schools were selected in successive stages and the number of students drawn from each school reflected its share of the total enrolment. This design reduces the risk that a single large or atypical school would dominate the results. Data collection relied on structured questionnaires administered with the help of class teachers, and the study proceeded with ethical clearance from the Chuka University Ethics Committee, a research permit from the National Commission for Science, Technology and Innovation, and authorisation from the County Director of Education for Machakos County.

The analytical centrepiece of the study is a Poisson regression model, a statistical technique chosen because the outcome of interest, counts of participation, is a non-negative integer rather than a continuous measure. Where ordinary linear regression assumes a normally distributed dependent variable, Poisson regression models count data directly, estimating how each explanatory factor changes the expected rate of participation. The researchers screened their candidate predictors for multicollinearity before inclusion, a standard safeguard against the distorted estimates that arise when explanatory variables overlap too heavily. The model was estimated separately for girls and boys, an approach that allowed the team to test not just whether participation levels differed by gender, but whether the underlying economic machinery of participation itself worked differently for each group.

The headline result is that it does. For girls, the strongest and most counterintuitive finding concerned parents’ disposable income: rather than boosting participation, higher disposable income was significantly associated with reduced participation in classroom activities. All other variables in the girls’ model, including the cost of education and family structure, showed no statistically significant effects. For boys, the picture looked markedly different. The cost of education emerged as a significant brake, with rising educational costs reducing the likelihood of participation, while two factors pushed in the opposite direction: students’ expectations of future employment and the number of siblings in the household both significantly increased boys’ participation. The remaining variables in the boys’ model, including parents’ disposable income, showed no significant influence. These gendered patterns held even as overall participation levels themselves also differed between girls and boys.

The counterintuitive income effect for girls invites careful interpretation. In low-resource households, disposable income may flow toward priorities that compete with girls’ schoolwork, or it may correlate with household arrangements in which daughters shoulder more domestic labour regardless of financial comfort. It is also possible that the relationship reflects how income interacts with expectations: families with more resources may channel girls toward different futures, or simply express their support in ways that do not translate into classroom assertiveness. The authors are careful not to overclaim a mechanism, but the practical implication is clear. Financial capacity alone does not guarantee that girls will find their voice in class, and programmes that hand money to households without attending to gender norms inside them may leave the participation gap intact.

The findings for boys are more intuitively legible. When school feels expensive, participation falls, consistent with a long body of evidence showing that educational costs, even in nominally free systems, filter down into engagement as families weigh marginal spending against competing needs. Kenya’s system of day secondary schools nominally benefits from public subsidies, yet levies, materials, uniforms and transport continue to strain household budgets in semi-arid districts like Mwala. The positive effect of sibling count on boys’ participation is intriguing and may reflect peer-like dynamics at home, shared study habits, or the confidence that comes from growing up in a crowded, verbally competitive household. Meanwhile, boys who expect employment after school participate more, exactly as human capital theory would predict: when education is perceived as an investment with a payoff, students invest effort in it.

That theoretical framing runs throughout the paper. The authors situate their work in the human capital tradition that views education as an investment whose returns justify present costs, and in expectancy-value accounts of motivation, which hold that students engage when they both value the outcome and expect to succeed in attaining it. Previous studies, from active-learning classrooms in higher education to analyses of gendered participation in online discussions, have documented persistent gender gaps in who speaks and how often, typically attributing them to classroom demography, confidence and social roles. What this study adds is an economic layer grounded in one of the contexts where those gaps matter most: resource-constrained rural secondary schools where household poverty, family size and labour expectations press directly on adolescent learners.

The policy implications the authors draw are correspondingly targeted. They call for gender-responsive interventions rather than blanket programmes, arguing that what lifts boys’ engagement, such as reducing the visible costs of schooling, may not address the specific barriers facing girls, and vice versa. They point toward strengthened school-based support systems, including guidance and counselling structures that can help students connect classroom effort to credible career pathways, and they urge that career guidance be enhanced so that employment expectations, which appear to energise participation, are realistic and widely shared across genders. For policymakers in Kenya’s Ministry of Education and for the international organisations tracking progress toward the Sustainable Development Goal on quality education, the study offers granular evidence that engagement, not merely enrolment, is where gender inequality now does much of its damage.

The research also speaks to a broader shift in education economics across the developing world. With primary and secondary enrolment having expanded dramatically across East Africa in recent decades, attention has moved from getting children into school to measuring what happens once they are there. Studies of this kind, which treat participation as a countable, modelable behaviour with identifiable economic determinants, are part of a growing toolkit for that task. The authors acknowledge the limits of a cross-sectional design, which captures association rather than proven causation and reflects a single sub-county at a single moment. Data will be made available on request, and the team is transparent that self-reported participation carries its own measurement challenges. Yet within those limits, the message is striking and portable well beyond Mwala: the same classroom contains students whose willingness to engage is being pulled by different economic strings, and any serious effort to close gender gaps in learning must read those strings separately, one group of learners at a time.

The study’s setting matters for interpreting its results. Mwala Sub-County lies in a semi-arid agroecological zone where household incomes depend heavily on rain-fed agriculture, making family finances vulnerable to seasonal shocks that can ripple directly into schooling decisions. Public day secondary schools serve the majority of students in such areas precisely because they allow learners to remain at home, but that arrangement also means classroom engagement is tightly coupled to what happens within the household each evening, from chores to paid work.

Methodologically, the choice to model participation counts separately by gender reflects a growing recognition in education research that pooled samples can mask offsetting effects. A factor that suppresses engagement among girls while boosting it among boys might appear insignificant in a combined model, hiding precisely the dynamics that gender-responsive policy needs to target. The multistage, proportionate sampling of 320 Form Three students strengthens the representativeness of the estimates within the sub-county, though the authors note that self-reported participation and the cross-sectional design limit causal claims.

The emphasis on employment expectations also aligns with expectancy-value frameworks of motivation, which hold that effort follows from believing both that success is possible and that it leads to valued outcomes. In contexts where visible pathways from school to work are unevenly distributed by gender, career guidance becomes not an accessory but a core lever for equitable engagement.

Subject of Research: Gendered socioeconomic determinants of student participation in classroom activities in Kenyan public day secondary schools

Article Title: Gendered determinants of students’ perceived participation in classroom activities: evidence from public day secondary schools in South Eastern Kenya

Article References: Mbithi, A. W., Mbaka, P. K., Mbaka, K. K., & Saoke, V. O. (2026). Gendered determinants of students’ perceived participation in classroom activities: evidence from public day secondary schools in South Eastern Kenya. SN Social Sciences, 6(9), Article 423. https://doi.org/10.1007/s43545-026-01720-1

Image Credits: AI Generated

DOI: 10.1007/s43545-026-01720-1

Keywords: gender differences, classroom participation, Poisson regression, cost of education, employment expectations, Kenya, secondary education, socioeconomic status, girls' education, educational equity, human capital, school engagement