Coverage bias: the impact of eligibility constraints on mobile phone-based sampling and data collection

Abstract

This paper uses Individual Deprivation Measure data from Indonesia and South Africa to demonstrate the effects of coverage bias associated with mobile phone-based sampling and data collection approaches that restrict sampling frames to those who own or have access to a mobile phone - a increasingly common method. Analysis of this data demonstrates substantial differences across multiple living standards indicators and demographic characteristics between those who own a mobile phone, individuals with access to a phone and those who neither own nor access mobile phones. This paper highlights the implications of such sampling approaches, and calls for the explicit consideration of the limitations of inferences that can be drawn from qualitative and quantitative analyses because of these substantive differences.