Sources of error in mobile survey data collection

The proliferation of mobile technologies in the general population offers new opportunities for survey research, but also introduces new sources of error to the data collection process. This thesis studies two potential sources of error in mobile survey data collection: measurement error and nonresponse. Chapter 1 examines how the diagonal screen size of a mobile device affects measurement error. Using data from a non-mobile-optimised web survey, I compare data quality between screen size groups. Results suggest that data quality mainly differs between small smartphones with a screen size of below 4.0 inches and larger mobile devices. Respondents using small smartphones are more likely to break off during the survey, to provide shorter answers to open-ended questions, and to select fewer items in check-all-that-apply questions than respondents using devices with larger screens. Due to the portability of mobile devices, mobile web respondents are more likely to be in distracting environments where other people are present. Chapter 2 explores how distractions during web survey completion influence measurement error. I conducted a laboratory experiment where participants were randomly assigned to devices (PC or tablet) and to one of three distraction conditions (presence of other people who have a loud conversation, presence of music, or no distraction). Although respondents felt more distracted in the two distraction conditions, I did not find significant effects of distraction on data quality. Chapter 3 investigates correlates of nonresponse to data collection using mobile technologies. We asked members of a probability household panel about their willingness to participate in various data collection tasks on their mobile device. We find that willingness varies considerably by the type of activity involved, to some extent by device, and by respondent: those who report higher security concerns and who use their device less intensively are less willing to participate in mobile data collection.