DATASETS
Social Connectedness Index
This index measures the strength of connectedness between two geographic areas as represented by Facebook friendship ties. SCI can reveal important insights about economics, social mobility, and health.
Download from HDX
Explore Africapolis SCI visualization
Key Features
Friendship data
SCI offers a new type of data to the research and non-profit community, measuring the frequency and density of friendship and social ties around the world.Global reach
With over three billion people using Meta services, this index provides the first comprehensive measure of social networks at an international level.Privacy
SCI publishes a number to show the probability of friendship between two areas, but no information on individual friendships. We use sampling, normalization, and differential privacy to protect privacy.
Methodology
Step 1: Assign people to geographic areas
We assign people to geographic areas based on their information and activity on Facebook, such as the stated city on their Facebook profile.Step 2: Calculate connectedness
We count the total number of friendship links of Facebook users between the geographic areas. We scale this value between 1 and one billion.Step 3: Drop small counts and add noise
We remove all locations with a low number of observations and add random noise to the number of friendships between each set of locations to reduce the likelihood that someone could be re-identified.Step 4: Final sampling
The final SCI is the average scale of friendship ties across 10 random draws from 99% of active Facebook users to further protect privacy.RESOURCES
Using the Social Connectedness Index
This index can be downloaded from the Humanitarian Data Exchange.
Download from the Humanitarian Data Exchange