Who Gets to Log On? Caste and Gender Shape Digital Learning Gaps in Nepal’s Universities

When the COVID-19 pandemic forced universities worldwide to move their classrooms online, the promise of digital learning collided with an uncomfortable reality: not every student could get online. In Nepal, where higher education institutions scrambled to adopt virtual teaching almost overnight, the question of who actually owns a device, who has a working internet connection, and who is left behind has taken on new urgency. A new study of master’s students at Tribhuvan University, published in Discover Education, offers one of the most detailed statistical portraits yet of basic digital access in a resource-constrained South Asian university system, and its findings reveal that the digital divide in Nepal’s capital is not a matter of simple connectivity but a mirror of deeper social stratification.

The research team, led by Yadu Ram Upreti of Tribhuvan University’s Central Department of Education together with colleagues at Tribhuvan University and Kathmandu University School of Education, surveyed 338 students enrolled in master’s in education programs across seven campuses in the Kathmandu Valley. The campuses included the university’s Central Department, two constituent campuses, and randomly selected community and private campuses. Students completed a structured, self-administered questionnaire covering their socio-demographic backgrounds, their educational circumstances, and their access to digital learning resources. The instrument was validated through expert consultation and pre-tested on 37 students, with reliability coefficients of 0.878 for the digital resources block and 0.888 for the digital competency block, indicating strong internal consistency.

The headline numbers are, on the surface, encouraging. Nearly all surveyed students, 96.2 percent, reported access to digital learning hardware, and 97.3 percent had internet access at home. Smartphones dominated the device landscape, used by 82.2 percent of students, followed by laptops at 48.5 percent, desktop computers at 11 percent, and tablets at just 7.4 percent. Wi-Fi was the most common home connection at 76.6 percent, with 26.4 percent relying on mobile data packages and only 10.9 percent using wired ADSL lines. But the study also uncovered a striking institutional gap: 45.6 percent of students reported having no internet access on their own campuses, meaning that nearly half of these future educators were studying in buildings without the connectivity their coursework increasingly demands.

Beneath the averages, however, the data tell a sharper story. Using chi-square tests of association followed by binomial logistic regression, the researchers found that gender and caste or ethnicity were significantly associated with access to digital hardware. Female students reported higher hardware access than male students, 97.8 percent versus 92.8 percent, a difference the authors link to rising female enrollment in higher education. More consequentially, students from Nepal’s historically advantaged Chhetri and Brahmin groups reported 97.7 percent hardware access compared with 93.1 percent among students from other ethnic communities, including Janajati, Madhesi, and Dalit populations. In the multivariate model, ethnicity remained a significant predictor: non-Chhetri/Brahmin students had roughly a third of the odds of hardware access compared with their Chhetri/Brahmin peers, with an adjusted odds ratio of 0.291.

The authors are careful to interpret this ethnic gradient not as an effect of identity itself but as an expression of structural inequality. In Nepal, caste and ethnicity have long structured access to land, secure employment, urban residence, and schooling. Because acquiring and maintaining digital devices depends on disposable household income, the observed gap most plausibly reflects the economic disparities that track these social categories. This reading aligns with the resource-based account of digital inequality developed by sociologist Jan van Dijk, whose framework the study adopts. Van Dijk’s theory treats access to technology as a sequential process, beginning with material access to devices and connectivity, and shaped by the unequal distribution of material, social, and cultural resources that are themselves stratified by social position. The study focuses squarely on this first-level divide, leaving skills and outcomes, the second and third levels, for future work.

Internet access at home followed a different pattern. Here, the type of institution, the student’s specialization subject, and gender all showed significant bivariate associations. Students specializing in language education reported the highest home internet access at 99.4 percent, compared with 97.2 percent for social sciences and just 90.7 percent for those in mathematics, science, and ICT specializations. In the regression model, specialization remained significant: math, science, and ICT students had markedly lower odds of home internet access than language students, with an adjusted odds ratio of 0.098. The researchers caution that this counterintuitive result, in which ICT-focused students are the least connected, likely reflects the social composition of disciplinary cohorts rather than the disciplines themselves. Differences in gender balance, urban or rural origin, and household economic background among those who enroll in each specialization are the more plausible drivers, and the small number of math, science, and ICT students in the sample, just 54, yields wide confidence intervals that warrant confirmation in larger studies.

The predominance of smartphones raises its own concerns. While mobile devices are portable and versatile, the authors note that reliance on them may limit students’ ability to engage with academic tasks that require more resource-intensive platforms, such as statistical software, long-form writing, or multimedia production. Prolonged smartphone use for academic work also carries documented health risks, including computer vision syndrome. Combined with the finding that many students depend on mobile data packages rather than high-speed Wi-Fi, the picture that emerges is one of access that is technically present but often fragile, expensive, or functionally inadequate for serious academic work. The study’s comparison with earlier Nepali research underscores how much has changed: a previous survey in Lumbini province found that only about half of public-campus students owned smartphones and two-thirds had home internet, whereas the Kathmandu Valley master’s students surveyed here fare considerably better, likely because they are more senior, often employed, and living in the country’s best-connected region.

The study’s limitations are candidly acknowledged. Data came from a single university in one urban valley, respondents were recruited by convenience sampling from students present on campus on survey day, and the cross-sectional design can identify associations but not causes. Because absent students, who may face the greatest barriers, were not surveyed, the reported access levels may be somewhat overestimated. The binary yes-or-no measures also captured only basic availability, not the quality, stability, or affordability of connectivity, nor students’ digital skills or how effectively they use what they have. The findings should therefore be read as indicative of regularly attending master’s students at the sampled campuses rather than as a national portrait of Nepali higher education.

Even so, the implications are concrete. The persistence of a first-level access divide patterned by caste, ethnicity, and economic background suggests that policy should move beyond generic calls for digitalization toward equity-oriented measures such as subsidized device schemes and affordable connectivity targeted at historically disadvantaged communities. The finding that nearly half of students lack campus internet points to an immediate, actionable priority: reliable, free, campus-wide connectivity. And because access alone does not guarantee meaningful use, the authors argue that infrastructure investment must be paired with sustained digital literacy training and discipline-specific support. Viewed through the digital divide framework, the deeper worry is that without deliberate intervention, the expansion of digital learning in Nepal will not narrow existing inequalities but convert them into a new, digital form of disadvantage, one in which the students most in need of flexible, connected education are precisely the ones least equipped to receive it.

Subject of Research: Digital learning resource access and digital divide among higher education students in post-pandemic Nepal

Article Title: Factors associated with access to digital learning resources among higher education students in post-pandemic Nepal

Article References: Factors associated with access to digital learning resources among higher education students in post-pandemic Nepal. (n.d.). https://doi.org/10.1007/s44217-026-02321-z

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02321-z

Keywords: digital divide, digital learning resources, higher education, Nepal, Tribhuvan University, COVID-19, ethnicity, gender inequality, internet access, smartphones, educational equity, survey research