Scientometric analyses of digital inequity in education: problems and solutions

The study addresses the growing concern of digital inequity in education, defined as unequal distribution of digital resources, opportunities, skills training, and development among diverse groups. The authors underscore the significance of researching digital inequity because it affects educational outcomes, learning environments, policy, and the design of educational technologies. Identified research gaps include a dominance of macro-level studies over micro-level focus on individuals; limited exploration of links between digital inequity and social or personal factors (gender, age, status, economics); a lack of longitudinal work; and scarce studies on concrete solutions. The study proposes to fill these gaps via a mixed scientometric and qualitative approach: quantitatively mapping trends and actors using VOSviewer and qualitatively synthesizing themes with NVivo. The research questions target yearly trends; top authors, sources, organizations, and countries; top research topics; and an in-depth categorization of digital inequities (access, usage/skills, quality of digital experience, information/content, economic, social exclusion, and gender-based).

The literature review outlines multiple dimensions of digital inequity in education. Access inequity stems from socioeconomic disadvantage, geography (rural/remote), and inadequate infrastructure, restricting devices and connectivity and dampening motivation and performance. Usage and skill gaps persist even with access, influenced by age, exposure, socioeconomic status, and educational background, impacting learners' ability to engage with digital tools and content. The quality of digital experience varies by device quality, internet speed, usability, and accessibility, shaping engagement and widening disparities. Information and content inequity arises from unequal distribution of quality resources, language barriers, and cultural bias, which can marginalize certain groups. Economic inequity interacts with digitalization, as affordability of devices and services constrains participation and can fuel cycles of disadvantage. Social exclusion in digital contexts limits interaction, community participation, and psychological well-being. Gender-based inequity is influenced by stereotypes and social roles that reduce women's engagement and skill development in digital domains. The review also traces topic evolution: early emphasis on technology use differences, shifting to equity concerns and structural issues (policy, funding, infrastructure) and socio-cultural factors (age, gender, race, SES). It motivates a scientometric lens to identify leading contributors and topics, noting prior bibliometrics were limited and often superficial.

The study employs a mixed scientometric and qualitative design using VOSviewer and NVivo, guided by PRISMA filtering. Data sources and retrieval: On May 30, 2024, 398 records were retrieved from Web of Science (All Databases; query: "digital inequit*" OR "digital inequalit*" (Topic) AND educat* (Topic); excluding Preprint Citation Index). Publication types included Article (323), Dissertation Thesis (35), Meeting (25), Early Access (20), Review Article (17), Awarded Grant (8), Editorial Material (4), Abstract (1), and Book (1). PRISMA screening: 29 duplicates and 12 automation-flagged ineligible records were removed, leaving 357 for screening; 164 were out-of-scope. Of 193 assessed for eligibility, exclusions included lower quality based on AERA standards (56), small sample sizes (58), and unconvincing results (46). A total of 35 studies were included for qualitative synthesis. Additional bibliometric datasets: For trend analysis, on March 28, 2024, a search for "Digital inequ*" (Topic) AND educat* (Topic) in Web of Science (excluding Preprint Citation Index) yielded 382 publications (time span 1637–2024), 7,592 citing articles (7,385 without self-citations), 9,444 cited times (8,981 without self-citations), average 24.72 citations per item, H-index 46. For mapping top authors/sources/organizations/countries, on March 28, 2024, 312 results from Web of Science Core Collection for "digit* inequ*" (Topic) AND educat* (Topic) were analyzed in VOSviewer. VOSviewer analyses: Co-authorship (units: authors, organizations, countries) and co-citation (cited sources) with full counting were conducted to identify top contributors by citation and link strengths. Co-occurrence analysis of author keywords set a threshold of at least two occurrences; 182 of 956 keywords met the threshold. NVivo analyses: The 35 included studies were imported for qualitative coding. NVivo supported auto-coding, thematic classification, text retrieval, word frequency analysis (e.g., generating a word cloud of the 2,000 most frequent words with minimum length 5), and a word tree based on exact matches of "digital inequity." Visualization (word cloud/tree) aided pattern discovery across themes.

Trends: Publications on digital inequity in education began growing around 2002, fluctuated until a rise in 2008, dipped in 2009, then increased, surging after 2018 and peaking in 2022, with a decline in 2023 (2024 incomplete at time of data collection). Across 382 items, there were 7,592 citing articles (7,385 without self-citations), 9,444 citations (8,981 without self-citations), average 24.72 citations per item, H-index 46. Top contributors (co-authorship/co-citation): Cited authors included Hargittai (1,636 citations), Van Deursen (889), Van Dijk (715), Cotten (157), Jansen (91), Van Der Zeeuw (91), Addeo (86), Ragnedda (86), Ruiu (86), Dodel (82). Top cited sources included New Media & Society (450 citations; 4,064 links), Information, Communication & Society (377; 3,529), Computers in Human Behavior (286; 2,513), Computers & Education (193; 1,433), Poetics (118; 1,371), Telematics and Informatics (112; 1,265), Communication Research (107; 1,197), Social Science Computer Review (106; 1,200), Telecommunications Policy (99; 805). Leading organizations by citations: Northwestern University (1,620), University of Twente (907), University of Zurich (368), Zhejiang University (244), Michigan State University (231), University of Amsterdam (198), University of California, Berkeley (110), University of Oxford (100), The University of British Columbia (93), Northumbria University (91). Leading countries by citations: USA (3,012), Netherlands (1,263), Canada (977), Germany (531), China (414), England (394), Switzerland (369), Italy (284), Spain (242), Russia (41). Top research topics (keyword co-occurrence): digital divide, digital inequality, digital inequalities, COVID-19, education, digital skills, digital inclusion, higher education, internet, internet use. NVivo word frequency indicated focal terms such as learning (3,171), education (2,235), technology (2,054), digital (1,963), students (1,836), online (1,532), access (970), teaching (794), school (786), social (616), equity (602), university (572), support (528). Qualitative synthesis categorized digital inequity into seven dimensions: access inequity; usage and skill gaps; quality of digital experience; information and content inequity; economic inequity in the digital age; social exclusion in the digital sphere; and gender-based digital inequity. Proposed solution strategies included: infrastructure expansion (broadband, devices) especially in remote areas; integrating digital tools into pedagogy and training teachers; project-based, hands-on student opportunities; partnerships with companies/communities to update resources; user-centered, accessible design with continuous evaluation; inclusive and diverse content addressing language and cultural bias; financial supports (devices, connectivity, scholarships); digital literacy programs; inclusive teaching practices and community-building online; and addressing gender norms via gender-responsive design, pedagogy, and literacy initiatives.

The findings address the research questions by offering a comprehensive map of the field’s evolution, its key contributors and outlets, and the dominant research themes, while qualitatively unpacking the multifaceted nature of digital inequity. Scientometric analyses delineate who and what drives the scholarship, guiding future collaborations, funding, and policy engagement. The qualitative synthesis connects structural, social, and cultural mechanisms with educational processes and outcomes, articulating actionable levers for change. The study’s contributions include: a systematic, data-driven overview of the state of research; integration of quantitative bibliometrics with qualitative thematic analysis for a fuller picture; and a critical appraisal of feasible solution pathways grounded in empirical and theoretical literature. These insights inform policymakers, instructional designers, and educators in designing inclusive digital learning ecosystems and interventions, and they contribute to refining theoretical frameworks around digital inequity by incorporating interacting factors such as equipment, autonomy, skills, social support, and purposes of use.

The study identifies pressing problems of digital inequity in education across access, skills, experience quality, information/content, economic factors, social exclusion, and gender dynamics, and proposes a suite of solutions spanning policy, technology, and pedagogy. A comprehensive approach—addressing cultural norms, investing in inclusive infrastructure and design, and boosting digital literacy—can help balance inequities and promote an inclusive education system. Future research should emphasize cross-cultural and cross-contextual comparisons, longitudinal designs to trace trajectories and impacts over time, mixed-methods integration to connect scope with lived experience, and normative, policy-oriented work to translate evidence into practical interventions and frameworks.

The mixed-methods approach may not capture all relevant themes due to the complexity of digital inequity. Hidden socio-cultural determinants may be under-identified. Generalizability is constrained by the limited number of included studies. Coverage limitations in available library resources mean some relevant literature may have been missed, introducing potential bias.