Visualizing Changes In The World’s Access To Internet


Charts & Graphs, Lab Reports, Visualization

Introduction

The internet has undoubtedly changed the everyday lives of many. While the earliest internet prototype, ARPANET, can be traced back to the late 1960s, the internet as we know it was born in the early 1980s. Over 90% of US adults use the internet. However, 4 billion people across the world are currently active users, which accounts for just 59% of the global population.

While one can estimate that the world’s overall access has probably increased over the most recent decade, I began this project interested in discovering how access might have disproportionately increased around the world. Using a dataset and information visualization software, I explored just how much life has changed in the decade between 1999 and 2019.

Screenshot of one of my Tableau visualizations.

Comparative Analysis

In constructing my visualizations, the following two struck me as meaningful. The first being the Share of the population using the internet. The overall message is a general increase in users. The overlapping lines create a striking message. However, I realized I was more interested in the percentage increases on more of a macro level.

The second visualization was a timelapse on a similar tech topic. It influenced my choice to incorporate a “dark mode” into my map and use the density of increases to shed light on areas where access has increased the most. Psychologically, I felt that using blue or cool-tones instead of rainbow would be an appropriate association to computers’ use of blue light.

Materials

After browsing datasets related to technology and the internet, I landed on The World Bank, which provided access to data on Individuals using the internet sourced from International Telecommunication Union’s Database.

OpenRefine was then used to clean up the data, and finally, Tableau was used to create several visualization worksheets which were then combined into a dashboard and posted on Tableau Public.

Methods

Data before refinement.
Data after transposing and filtering.

After saving as a .csv in Numbers, I used Open Refine to Transpose the years from Columns into Rows as the original data had dozens of columns spanning over 30 years. After playing around with this data in Tableau, I decided to go back into Numbers and filter out all years except 2019 and 1999 to simplify any calculations, since I knew I wanted to focus on changes within the most recent decade (2020 data was unavailable).

The .csv was then imported into Tableau where I began creating several visualizations. I started with a symbol map and used a Quick Table Calculation to calculate the Difference by Year (this took a bit to figure out). Using the size of the circle to plot the % difference seemed to give the illusion that circle size represented current access. Since I wanted to focus on the increase in access, I switch to Density marks. This provided a radial gradient to demonstrate the % increase.

The original version (above).
The new cool-toned gradients provide an illusion of glowing, with the brightest areas being those that experienced the greatest increase in internet access between 1999-2019 (above).
Only the maximum and minimum were assigned a color.

While a map felt appropriate for presenting countries’ increases, I chose to provide a different visualization to present 2019 % of users for reference. The dataset already included this information by region, so I started by using a packed bubble visualization with the specific Regions filtered. I then switched to horizontal bars to allow users to more easily make comparisons or estimate ratios. In my final visualization, I hid the Region key since all countries (except North America and Sup-Saharan Africa) were colored grey, to emphasize the Min and Max (see right).

The dataset also included information related to internet access by income level. This was the most interesting data to me. I tried out different visualizations before deciding on side-by-side circles. Since I was using only two years (1999 and 2019), this felt better suited than lines, which might oversimplify the year-to-year differences.

The columns were Region and Years (set to discrete), with the rows being the % for each year.

The circles’ position on the Y-axis and circle size represent the % increase at each income level.

The three visualizations were combined into a final dashboard for ease of view:

Interpretation

Screenshot of the final Tableau Public dashboard.

Internet usage is clearly increasing- but the density map provides insight into which regions have seen the swiftest changes. The Middle East, Central Europe, and the Caribbean have had the biggest explosion of access in the past decade. Access in the United States increased 51.42%, however, countries like small Aruba have seen over a 90% increase.

North America currently has the highest rate of access at 87.65%, while Sub-Saharan Africa is at 25.08%.

The most fascinating results were from the Change in access by income graph:

I found that while low-income populations are at 15.30% access in 2019, this amount was just less than 7% of the high-income populations’ access back in 1999. Meaning 10 years ago, higher-income people had access even more access than lower-income people today.

Reflection

What would a world be like where 100% of the population had access to the internet? While I focused on 1999 and 2019, it would be interesting to calculate the total rate of increase by country to forecast the potential future year of full access saturation.

How far back can we go? There was also data provided before 1999, however, it was very limited in the number of countries. It would be interesting to source this data elsewhere to create a time series with this information.

What factors contribute to increasing access? Uncovering another data set with policy and funding changes around the world could help to uncover valuable causation.