Eliciting User Experience Information in Early Design Phases. The CARE Approach to In-Vehicle UX

Doctoral thesis, 2018

Experience-rich input in early phases of a design process can offer valuable information and inspiration to designers. However, there are methodological challenges linked with efforts to understand future user experiences. Experience encompasses multi-layered and tacit data, such as emotions and value, that are important for commercial success but are difficult to elicit from users for existing products, and even more so for concepts in early design phases. At early design phases, the inevitably incomplete representations of product and use context influences the outcomes. It is typically easier to elicit usability-related aspects, meaning that other aspects of experience may be insufficiently addressed. The contribution of this thesis is an approach for eliciting rich user experience (UX) data in early design phases, building on six studies. This thesis employs in-vehicle user experience as a study case, but results are however presented on a methodological level that can also be of use to other interactive products. The overall research questions are: What signifies in-vehicle UX? How can UX data be elicited for input to novel in-vehicle concepts in early design phases?

Firstly, the analysis phase of the design process was addressed, where a multi-method approach was employed to study current in-vehicle UX. UX is an umbrella term that has proven difficult to describe and conceptualise in studies. Therefore, the aim of the first study was to better understand what signifies the specific case of in-vehicle UX. Secondly, how to approach and understand user expectations on future autonomous cars was address in the two following studies, in order to address prospective research of novel systems. A method addressing research on user expectations was developed – Setting the Stage for Autonomous Cars. Thirdly, ideation was addressed in a series of workshops, containing generative and creative efforts for ideating future interactive in-vehicle systems. Methods such as enactment, small-scale scenarios, Wizard of Oz, a lo-fi driving simulator and the developed Setting the Stage for Autonomous Cars method were used. The final studies address concept evaluation, and comparatively explore the effects of choosing different product representations (storyboard and interactive prototype) and study contexts (Virtual Reality and in the field) in early UX evaluation.

Based on the outcomes of the studies, an approach is proposed – the CARE approach – for enabling richer and more in-depth UX data in early design phases. This approach suggests that there is a need to Contextualise the researched experience (conveying the intended use situation and sentisising the participants to experience), enabling the participant to Act (enabling interaction even at the stages of very lo-fi concepts), supporting Reflection on the experience (enhanced by generative elements in the methods, such as drawing concepts and enacting use) and enabling the participant to Express the experience (in more ways than by just relying on words). Furthermore, the thesis presents findings regarding what signifies in-vehicle UX, for example whole-body, multi-sensory interactions, the importance of the temporal stage of use, the social and multi-device context, and the changing relationship between user and car with increased automation. The results emphasise the importance of addressing the multisensory use situation in each design phase and for participants to express experiences, not only in words but also through enactment and generative techniques.