Eye movement recordings in natural settings

In this chapter we consider why it is important to study eye movements in natural, real world settings and the practical considerations that need to be taken into account when collecting eye movement data in these situations. Conducting a study in a real world environment poses very different challenges than those present in laboratory-based settings. Variables are hard to control and this restricts the kinds of paradigms that can be employed. Thus, careful consideration of the research question and whether it is appropriate for real world study is a necessary first step. Mobile eye trackers are often the most obvious choice for studying eye movements in real world settings. There exists a range of mobile eye tracking systems available to buy, and researchers also build their own mobile eye trackers to solve particular issues. Selecting the right system is an important consideration for mobile eye tracking and in this chapter we will highlight some of the key choices that the researcher should consider when buying or building a mobile eye tracker. Care must be taken to ensure that the data collected are sufficiently reliable to address the research question. While the principles of eye tracking - how we detect features of the eye and use these to estimate where someone is looking - are very similar for all eye trackers, the challenges faced when collecting data in natural settings are far greater than those that face eye trackers in a controlled laboratory setting. We will consider the key threats that real world settings pose for collecting reliable eye tracking data in the hope that we not only raise awareness of these issues but also offer some suggestions for how these can be minimised or at least identified in the data. For those considering whether to run an eye tracking study in the real world, we hope that this chapter will help you as you consider your research questions and choices of mobile eye tracker. For those already committed to studying eye movement behaviour in a real world setting, we hope that our discussion of the problems and possible solutions to the challenges faced in such settings will help minimise data loss from the common challenges that we face when conducting mobile eye tracking studies.