Sensor-supported unsupervised observation techniques for filed studies

We live in a mobile era, and interactive, ubiquitous applications and services surround us every-where. To design and develop targeted and successful applications for these environments, a detailed understanding of the users and their mobile contexts is needed. Unfortunately, existing observation methods are mostly unsuited for these highly dynamic settings, because of their low scalability, limited situatedness, and high obtrusiveness. This thesis investigates to what extent the sensing and computation capabilities of smart phones, which are a key component in many ubiquitous systems, can be used to improve three of today’s most popular unsupervised observation techniques: logging, the Experience Sampling Method (ESM), and diary studies. It is demonstrated that the information gain of logging can be increased, if the statistical analysis is enriched with static, environmental information. Further, it is shown that situation-aware self-reporting techniques, such as ESM, can trigger inquiries in opportune moments and, therefore, are perceived as less obtrusive and are more likely to be answered. Finally, the thesis illustrates that properly prepared and presented in-situ information, e. g., multimedia diary entries, can valuably inspire and improve post-hoc observations, like follow-up interviews.