Internet Access and Education in South America

Problem

Connectivity is widely assumed to improve education, yet South America's digital divide is still wide. For the BUas SDG Hub, I tested that assumption: how closely has internet access tracked educational completion from 2000 to 2023 across ten South American countries: Argentina, Brazil, Chile, Colombia, Ecuador, Guyana, Paraguay, Peru, Suriname and Uruguay?

What I did

I merged two public datasets never designed to fit together, World Bank internet-use figures and UN SDG completion rates, reshaping, cleaning and standardising them in Python. I measured the relationship with correlation, linear regression and year-over-year growth, then built a Power BI dashboard so non-technical stakeholders could explore it country by country.

Result

Internet access and completion move together, but access explains only 34% of the variation. The rest likely sits in factors such as GDP per capita and education spending, the next variables I'd add. The pooled figure was pulled down by level differences between countries (Suriname started far below the rest), so I made every country's own correlation visible: within each country, the relationship is much stronger (0.59–0.98).

What 0.58 looks like

Skills gained

• Multi-source data integration: reshaping and joining World Bank and UN SDG datasets on composite country–year keys in pandas
• Data quality control: handling missing values, removing outliers and standardising inconsistent units
• Statistical modelling: Pearson correlation, linear regression, R² and year-over-year growth analysis
• Dashboard design: building an interactive Power BI dashboard for non-technical stakeholders
• Analytical validation: testing aggregate results at country level to catch misleading averages