The impact of biases in mobile phone ownership on estimates of human mobility

Department of Engineering and Public Policy, Carnegie Mellon University, , 5000 Forbes Avenue, Pittsburgh, PA 15221, USA.
Journal of The Royal Society Interface (Impact Factor: 3.86). 01/2013; 10(81):20120986. DOI: 10.1098/rsif.2012.0986
Source: PubMed

ABSTRACT Mobile phone data are increasingly being used to quantify the movements of human populations for a wide range of social, scientific and public health research. However, making population-level inferences using these data is complicated by differential ownership of phones among different demographic groups that may exhibit variable mobility. Here, we quantify the effects of ownership bias on mobility estimates by coupling two data sources from the same country during the same time frame. We analyse mobility patterns from one of the largest mobile phone datasets studied, representing the daily movements of nearly 15 million individuals in Kenya over the course of a year. We couple this analysis with the results from a survey of socioeconomic status, mobile phone ownership and usage patterns across the country, providing regional estimates of population distributions of income, reported airtime expenditure and actual airtime expenditure across the country. We match the two data sources and show that mobility estimates are surprisingly robust to the substantial biases in phone ownership across different geographical and socioeconomic groups.

Download full-text


Available from: Abdisalan Noor, Jul 28, 2014
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The understanding of mass population movements has greatly advanced with the rapid spread of ubiquitous devices. Anonymized call detail records (CDRs) for mobile phones have enabled us to not only trace individual trajectories but also approximate activity patterns, including significant locations such as homes and workplaces. The majority of studies analyzing CDRs attempt to utilize the mobility patterns of anonymized crowds to improve transportation and public health. This is quite reasonable because CDRs can capture the movements of people at given times and places, whereas general statistics usually account for a population based on their locations of residence. However, it has also been pointed out that there are discrepancies between the movements of people as observed through CDRs and those of an entire population in a given area. This is because CDRs only represent device users. In fact, we can never learn about the population that is unobservable through CDRs only by analyzing CDRs. Therefore, this study attempts to provide clues to help us understand the whereabouts of the unobservable population by analyzing two months of the CDRs for 58 volunteers with mobile device service from a major telecommunications company in combination with field survey data from Dhaka. We surveyed the personal and household attributes of mobile users in relation to their calling behavior. The analysis results show that per mobile user observed in CDRs, there is an average of roughly 2.4 to 2.8 unobservable people. Their age groups and gender composition are also provided. We find that male and female users exhibit opposite trends in call locations according to the presence of children within the household. In addition, based on field observations, we find that the location and time distributions of small children follow some specific routines. Our findings contribute to the understanding of the whereabouts of the unobservable population, the majority of whom are children and are considered to be vulnerable to disasters or infectious diseases but are difficult to locate through CDRs alone.
    2015 IEEE International Conference on Pervasive Computing and Communications (PerCom2015), St. Louis, USA; 03/2015
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The recent development of telecommunication networks is producing an unprecedented wealth of information and, as a consequence, an increasing interest in analyzing such data both from telecoms and from other stakeholders' points of view. In particular, mobile phone datasets offer access to insights into urban dynamics and human activities at an unprecedented scale and level of detail, representing a huge opportunity for research and real-world applications. This article surveys the new ideas and techniques related to the use of telecommunication data for urban sensing. We outline the data that can be collected from telecommunication networks as well as their strengths and weaknesses with a particular focus on urban sensing. We survey existing filtering and processing techniques to extract insights from this data and summarize them to provide recommendations on which datasets and techniques to use for specific urban sensing applications. Finally, we discuss a number of challenges and open research areas currently being faced in this field. We strongly believe the material and recommendations presented here will become increasingly important as mobile phone network datasets are becoming more accessible to the research community.
    ACM Computing Surveys 01/2014; 47(2):1-20. DOI:10.1145/2655691 · 4.04 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: While the size of cities is known to play a fundamental role in social and economic life, its impact on the structure of the underlying social networks is not well understood. Here, by mapping society-wide communication networks to the urban areas of two European countries, we show that both the number of social contacts and the total communication intensity grow superlinearly with city population size according to well-defined scaling relations. In contrast, the average communication intensity between each pair of persons and, perhaps surprisingly, the probability that an individual's contacts are also connected with each other remain constant. These empirical results predict that interaction-based spreading processes on social networks significantly accelerate as cities get bigger. Our findings should provide a microscopic basis for understanding the pervasive superlinear increase of socioeconomic quantities with city size, that embraces inventions, crime or contagious diseases and generally applies to all urban systems.
    Journal of The Royal Society Interface 10/2012; DOI:10.1098/rsif.2013.0789 · 3.86 Impact Factor