Understanding individual travel patterns in African cities

This pioneering study combined face-to-face interviews, Big Data instruments and detailed socio-economic information on households, interviewed by the Measuring Living Standards in Cities (MLSC) survey.

The data was gathered from 300 randomly selected MLSC households, using sensor-embedded smartphones which tracked travel by time and location.

Attempting to side-step selection biases commonly associated with Big Data applications, our work provided insights into individual travel preferences and purposes, as well as economic costs of congestion.

Challenges

‘Big Data’, and its potential to revolutionise the way information science is conducted, has received a lot of hype in recent years. However, if this data is to be used for public policy, we must consider the selection bias inherent to the type of data generated by digital footprints. This is particularly true in low- and middle-income countries, where those who are most disadvantaged are also those who will have the smallest digital footprints.

This project sought to side-step this challenge by providing sensor-embedded mobile phones to a random sample of households in Tanzania’s capital city, Dar es Salaam. It looked at travel patterns and time usage, which are inherently difficult to collect accurate data using traditional survey methods.

Our approach

We developed survey instruments and planed the complex implementation of all data collection and data management activities. Households were carefully selected from the preceding MLSC survey and linked to three rounds of data collection to determine eligibility for the study, administer mobile phones with IDs linked to the survey data, and follow-up with phone calls inquiring about daily activities and travel patterns. Participants were called twice per week for six weeks to verify travel patterns indicated through the GPS tracking on their phones. These activities required extensive data management, as well as monitoring and quality control, all of which were ensured through our system. Given that this project was the first of its kind, this particular approach was adopted in order to adjust for challenges that arose during implementation.

Outcomes

The completion of this innovative survey provided high-quality, high-frequency data on travel patterns and use of time in Dar es Salaam. In addition, this study will help inform future studies of its kind in Tanzania, the wider region and around the world.