Representativeness of mobile phone location data for accurate measurement of population movement

Human travel is used as a proxy for a range of social behaviors including migration, economic engagement, social connections, and the transmission of infectious disease. The widespread adoption of mobile phones has greatly enhanced scientists' ability to collect information on patterns of travel. While mobile phone location data provides undeniably useful information on empirical patterns of movement, this data may be collected from a sample of individuals which is not representative of the broader population which it is intended to represent. This research will address whether the sample of individuals included in mobile phone location data is biased with respect to spatial, behavioral, or demographic characteristics, and whether the movement captured in these datasets is an accurate proxy of specific social behaviors. This research will focus in particular on the use of mobile phone location data for predicting infectious disease transmission.

Research Areas: human geography, demography, sociology

Aims of the research:
1. The principal aim is to understand how geographic, social, and economic factors influence movement
behaviour during the COVID-19 pandemic.
2. I will explore the representativeness of movement data, and whether certain demographic groups or
movement types are under or overrepresented. Importantly, this will inform methods for generalising the
movement of a sample of individuals to the behaviours of an entire population.
3. I will examine the prediction of movement behaviour during emergencies and how these predictions
can be improved with information on the socio-demographic characteristics of populations.
4. I will investigate how privacy preserving aggregation techniques (spatial and temporal) impact
movement indicators and the effect these indicators have on different socio-demographic groups.

Detailed Research Questions: I will use movement data to quantify the varied geographic, social, and
economic determinants of movement behaviour in the UK. As in my previous work, this research will be
specifically applied to understanding patterns of movement during public health emergencies, including
the COVID-19 pandemic and associated interventions. The research will also make predictions of
movement patterns and determine how demographic factors improve the performance of predictions of
human movements. Finally, the research will determine the influence of privacy preserving aggregation
and censoring methods on the interpretation of human movement indicators.