This project will examine mobile devices, specifically smartphones and tablet computers, as vehicles for survey data collection. The appeal of these devices for survey researchers is clear. Because they are lightweight and relatively inexpensive, they make it easier to collect data using such existing survey modes as computer-assisted personal interviewing. The research will examine three issues raised by use of such devices. First, the input methods that these devices permit (such as touchscreen interfaces) are relatively unfamiliar to many users and may create response problems. Although these interfaces are sometimes used on laptops, tablets and smartphones require them, making usability concerns more central. Second, the screens on tablets and smartphones are considerably smaller than those on laptop or desktop computers. Experiments on web surveys demonstrate the importance of "visual prominence." Any information that respondents need to use should be immediately visible to them without their having to perform any action (such as a mouse click) to make the information visible. Even the need for an eye movement may effectively render information invisible. Because of the small screens on mobile devices, it may be much harder to make all of the potentially useful information visible to respondents than it is with a laptop or desktop computer. The final issue is the perceived privacy of data collected on these devices. Respondents are willing to reveal sensitive information about themselves when a computer administers the questions, and web surveys seem to retain the advantages of earlier methods of computerized self-administration. But it is unclear whether respondents will display the same level of candor when the survey is administered over the Internet on a tablet computer or a smartphone. Two realistic field experiments and a usability study will examine these issues. Both experiments will be conducted in a single, face-to-face survey. The first experiment will compare laptop computers with tablets and smartphones and will examine the effects of both screen size and input method on breakoffs, missing data, completion times, and indicators of the quality of the responses. The second experiment will compare the same three data collection platforms as vehicles for collecting sensitive information. The experiment will ask respondents to assess the sensitivity of the questions, because item sensitivity may vary as a function of the device used to collect the data.
Surveys are a central tool for social scientists and policymakers in the United States, and survey research is a multi-billion dollar industry in the United States alone. Any set of technological advances, such as the widespread adoption of smartphones and tablet computers, is likely to have a major impact on how surveys are done. Although mobile devices will be widely used for surveys regardless of whether this research is done, the work will produce practical guidelines for using such devices to collect survey data and will alert survey researchers to some of the potential pitfalls of these devices.
