Building Collaborative Test Practices: Design Ethnography and WOz in Autonomous Driving Research Katalin Osz 1,2 , Annie Rydström 2 , Vaike Fors 1 , Sarah Pink 3 Robert Broström 2 1 School of Information Technology, Halmstad University, 30118, Halmstad, Sweden 2 Volvo Car Group, 40531, Gothenburg, Sweden 3 Faculty of Information Technology, Monash Art, Design and Architecture, Monash University, Melbourne, VIC 3800, Australia {katalin, osz}@hh.se Abstract. This article outlines a novel way of performing experimental “Wizard of Oz” (WOz) User Experience (UX) research that specifically targets driving in different levels of self-driving modes. The reasons for exploring the possibilities of combining experimental and ethnographic WOz-testing have been twofold. On the one hand, this mixed-method approach responds to a growing body of critique concerning how the WOz test is biased by the claim that it explores real-life behaviour in an experimental setting. On the other hand, our approach also meets the demands for innovative research methodologies that can contribute to deeper understandings of how to better evaluate and account for human expectations and experiences when automated technologies become integrated in everyday life contexts. This knowledge is inevitable for a broader understanding of the overall user experience and expectations of autonomous driving and, more specifically, building an interdisciplinary collaborative testing approach. Keywords: autonomous cars, user experience, design anthropology, future technology, mixed-method approach 1 Introduction In this article, we outline and demonstrate a new design research approach to creating insights about human expectations and experiences of a technology that does not yet fully exist - Autonomous Driving (AD). Significant advances in technology have made AD of vehicles a technological reality. Developing autonomous cars is technically challenging and to date the primary research focus regarding human behaviour in relation to AD cars has been on safety critical aspects, such as people’s ability to take over control from the automated car [1], [2], [3], [4], [5], mode awareness [6], overtrust [7, 8], or system transparency [9], with the intention of finding out how to provide the driver with optimal information through the user interfaces [10]. In the area of public and user acceptance of autonomous vehicles, a number of existing studies have focused on expectations through research into how to establish trust and mitigate resistance toward autonomous driving [11], [12], [13], Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 12-20 12