The index arrives at a moment when the stakes of connectivity could hardly be higher. The COVID-19 pandemic turned internet access from a convenience into a lifeline, underpinning remote learning, telemedicine, and e-government. Yet scholars have long recognized that the digital divide is no longer a simple binary between the connected and the unconnected. It is a layered phenomenon shaped by skills, cost, infrastructure quality, and social context. In low- and middle-income countries, high data prices and scarce devices keep people offline even where networks technically exist. Women are consistently less likely than men to own phones or use digital services. Rural areas lag behind cities. More than a billion people worldwide face accessibility barriers in digital platforms. In fragile states such as Somalia, all of these pressures are compounded by weak institutions, security concerns, and political instability that deter investment in infrastructure.

To measure this layered reality, the researchers surveyed 80 respondents drawn from Somalia’s telecommunications sector and academia in Mogadishu, a purposive sample of the stakeholders most engaged with the country’s ICT ecosystem. The sample size was computed with a standard statistical formula for a 95 percent confidence level and a 5 percent margin of error, and the questionnaire’s reliability was confirmed with Cronbach’s alpha of 0.82, well above the conventional 0.70 benchmark for internal consistency in social science research.

The technical heart of the study is the construction of the index itself. Because the original survey scales varied in measurement units, responses were rescaled into a uniform 0-to-1 range using min-max normalization. Negative indicators, such as perceived service cost, were inverted so that larger values always represent better inclusion conditions—a higher score means more affordable, not more expensive. The three dimensions were then combined with weights assigned according to theoretical relevance and empirical reliability from the digital inclusion literature: digital literacy received the highest weight of 0.40, reflecting consistent evidence that skills are the strongest determinant of meaningful participation beyond simple access, while affordability and infrastructure adequacy each received 0.30 as complementary structural enablers. A respondent’s DII is therefore the weighted sum of their normalized literacy, affordability, and infrastructure scores.

The team stress-tested this weighting scheme with a sensitivity analysis comparing the weighted index against an equally weighted alternative of 0.33 for each dimension. Subgroup rankings and overall interpretations remained largely unchanged, indicating the findings are robust to reasonable variations in weighting assumptions. Exploratory factor analysis and principal component analysis were deliberately avoided because the small sample and composite dimensions would render them statistically unreliable; the authors flag these techniques as validation priorities for future research using larger datasets.

The results paint a portrait of Somalia’s most digitally active population. Sixty percent of respondents were male and 40 percent female. Fifty-one percent were aged 18 to 24 and 33 percent were 25 to 34, with only 16 percent older than 35. Eighty-four percent held university degrees. Smartphones dominated internet access at 55 percent of responses, followed by laptops at 35 percent and desktops at just 10 percent—a mobile-first pattern typical of low- and middle-income countries, where cost constraints, unreliable electricity, and portability advantages have pushed stationary computing aside. Communication was the most frequent internet use, followed by online learning and information search, while e-business, job searching, and e-government services showed minimal engagement, suggesting digital platforms remain poorly embedded in Somalia’s economic and governance structures.

The index itself revealed a mean DII of 0.61 with a standard deviation of 0.12—moderate but uneven inclusion. The subgroup differences were striking. Respondents aged 35 and above recorded the lowest average score at 0.51, highlighting generational skill gaps. Males scored slightly higher than females, 0.61 versus 0.57, primarily reflecting greater laptop ownership among men, though this gap was narrower than typically observed in many low- and middle-income countries. The most dramatic variation emerged along educational lines: respondents with a master’s degree or higher achieved the highest score at 0.67, while those with only secondary education scored substantially lower at 0.48—a gradient the authors identify as the most significant predictor of digital inequality in the sample.

Inferential statistics sharpened the picture. A chi-square test confirmed a statistically significant association between education level and digital service use. Binary logistic regression showed that respondents with a bachelor’s degree were 2.4 times more likely to use e-government services than those with secondary education, with an odds ratio of 2.4 and a 95 percent confidence interval of 1.1 to 5.5, while master’s degree holders were 3.6 times more likely, with a confidence interval of 1.4 to 9.1. Ordinary least squares regression showed that each one-level increase in educational attainment was associated with a 0.07-point increase in DII score, with a standard error of 0.02 and a p-value of 0.002, after controlling for age and gender. Males scored marginally higher on the index and younger respondents tended to score higher as well, but neither effect remained statistically significant in the adjusted models. Education, in short, is the variable that matters.

When asked directly about barriers, respondents identified lack of digital literacy as the most significant factor, cited by 37 respondents or 29 percent, followed closely by high cost of service at 28 percent, lack of infrastructure at 24 percent, and political instability at 19 percent. Perceptions of the digital landscape itself were revealing: 81 percent recognized a gap between rural and urban areas, 69 percent believed political instability worsens the divide, and 58 percent agreed the government promotes digital inclusion, though 28 percent remained uncertain. The authors are careful to note that these findings reflect the perspectives of digitally engaged stakeholders in Mogadishu rather than a nationally representative survey, and should not be generalized to rural communities or marginalized populations where infrastructure and access differ substantially.

Even with those caveats, the study maps each barrier to an actionable solution and offers an integrated policy roadmap. The recommendations include community-based digital literacy programs targeting people with lower educational attainment, women, older adults, and rural populations; public-private partnerships to subsidize internet costs and launch low-cost smartphone initiatives; expanded broadband infrastructure beyond urban centers; stronger governance frameworks to encourage private investment; and e-government and e-business platforms designed according to universal accessibility principles. The proposed index, the authors argue, offers a practical framework for assessing multidimensional digital inclusion in fragile-state settings—one that future research should validate against nationally representative datasets and internationally recognized indicators from bodies such as the International Telecommunication Union and GSMA. If the first number from Somalia’s digital ecosystem is any guide, the path to inclusion runs not through cables and towers alone, but through classrooms.

Subject of Research: Science

Article Title: Development and validation of a multidimensional digital inequality index for assessing digital literacy, affordability, infrastructure, and socio-demographic determinants of digital inclusion

Article References: Ishaq, O. M., Islam, F., Alam, M. J., Mumin, A. R. O., Hossain, S. M., Karim, H. A., & Sarker, M. T. (2026). Development and validation of a multidimensional digital inequality index for assessing digital literacy, affordability, infrastructure, and socio-demographic determinants of digital inclusion. Discover Education, 5(1), Article 842. https://doi.org/10.1007/s44217-026-02077-6

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02077-6

Keywords: digital divide, digital inequality, Somalia, digital literacy, digital inclusion, ICT policy, education, fragile states, Development, validation, multidimensional, digital