Chapter 5 Errors of nonobservation: Non-response, including non-consent, non-compliance

5.1 Nonparticipation errors and biases

We deal with nonparticipation bias, that is systematic differences between those who participate in these studies and those who don’t, given coverage.

5.2 Consent

We show examples of consent processes in wearables, sensors, and apps studies; report the results of randomized experiments that identify reasons for nonconsent and study design features that increase consent rates. We discuss the ethical side of consent and the need for researchers to strike a balance between providing information for the consent to be informed in a way that is comprehensive to participants.

5.3 Noncompliance

We focus on continuing participation, reasons for noncompliance (both human and technical), design features that allow to increase compliance and enhance apps and wearables experience, such as providing feedback to participants.

5.4 How to increase consent and compliance

We provide practical advice on how to increase consent and compliance that is based on randomized experiments and can be implemented in different data collection settings.

5.5 References