Unobtrusive Observation of Cycling Tourists in the Wild

Introduction

User studies are an essential and important part of mobile Human Computer Interaction (HCI). They are typically conducted at the beginning or at the end of a design cycle of an interactive system to understand the context of the user or to evaluate a developed concept. There exists a variety of different in-situ or post-hoc study techniques, like shadowing or interviews. While most of these study techniques are suited for different use cases and they are widely applied in the field, there are scenarios where traditional methods perform insufficiently or fail.

We know from our own long lasting experience that cycling as a leisure activity during holidays (see Figure 1), is a representative scenario, which can hardly be investigated with traditional observation methods (Pielot, Poppinga, & Boll, 2009; Poppinga, Pielot, & Boll, 2009; Pielot, Poppinga, Heuten, & Boll, 2012). Tourists on bicycles often cycle for several hours and try to forget about everyday routines and problems to experience ultimate relaxation and simply enjoy the landscapes and surroundings. In-situ observation techniques, like shadowing, disturb the intimate feeling of cycling and can influence the participants in their natural behavior. In addition, the permanent involvement of experimenters is costly and limits the scalability of the study. Other in-situ techniques, like diary studies, don't rely on the experimenter, but require active and repetitive user feedback. Thus, they would require the cyclist to stop or interrupt the experience, which is not appropriate. Post-hoc study techniques, like interviews, involve the participant after the actual experience. Since the participants have to reflect on earlier experiences, their descriptions often lack details and situatedness as the participants tend to forget or do not report details. The community is aware of these methodological limitations and proposes to investigate better field observation methods (Kjeldskov & Graham, 2003; Kjeldskov, Cheverst, de Sá, Jones, & Murray-Smith, 2012; Kjeldskov & Paay, 2012).

Figure 1.

Cycling as a leisure activity is often an intimate experience. Traditional observation techniques, like shadowing or diary studies, often fail in this setting as they are too obtrusive or request the cyclist to interrupt his or her experience.

In the last few years, hybrid observation approaches were proposed and used, e.g., Eldrige, Lamming, Flynn (1993) and Kalnikaite, Sellen, Whittaker, and Kirk (2010). The idea behind hybrid observation is to use unobtrusive in-situ observation techniques, like logging, that do not require an experimenter for the actual observation. In a post-hoc setting the gathered in-situ data is then used to provide objective insights and thereby jog the memories of the study participants to obtain additional qualitative feedback. Existing work focused on the application of hybrid observation to lifelogging or in repeating day-to-day scenarios, where study participants mostly deal with established, reoccurring behaviors. Until now, the applicability and values of the approach in less familiar scenarios with privacy constraints, like the mentioned cycling scenario, remain unclear.

In this paper, we apply a hybrid observation approach to such a field study. We investigate how 11 tourists use a vibro-tactile bicycle orientation aid during their holidays. By doing so, we want to contribute to the understanding of the technique in non-everyday settings and field studies. In detail, we want to investigate and clarify on the following research questions:

• •

  Which in-situ data collection technique is least obtrusive and disturbing?

• •

  How does the hybrid observation approach affect the relation and communication between experimenters and participants?

• •

  How and which in-situ data is used to discuss which aspects in a post-hoc interview?