Page 81 - WHERE WE WORK - Schlegelmilch

Moving between places
to a context - at the same time (Marwick & boyd, 2011), hence collapsing the contexts. These collapses can be unintentional, which are called "collisions" (Davis & Jurgenson, 2014, p. 481) and are often seen as problematic with chaotic consequences (Davis & Jurgenson, 2014). There are also intentional context collapses, "collusions" (Davis & Jurgenson, 2014, p. 480), which have more positive consequences such as knowledge sharing. The ambiguity of norms in a place can also give rise to uncertainty (Van Gennep, 1960) and creativity (Sturdy et al., 2006). A departure point for a follow-up study could be to focus on the digital nomads’ boundary work by observing how they separate and, or, blur the contexts in their spatial environment and what the implications are for their work.
2.5.2 Limitations and future research directions
The current study is subject to several limitations that, it should be noted, are also opportunities for future research. While we conducted a good number of interviews, the data is cross-sectional and provides only current and retrospective data about how digital nomads create their work environment. Such verbally recollected data has been found to be vulnerable to the recall effect (Golden, 1992). To counteract this and improve recall, we asked about what happened in a particular instance, where they were and what tools they used. We believe that by asking the respondents for such an amount of detail (e.g., the location and people involved), they were only able to recall these events if they had taken place. Furthermore, we complemented the interviews by conducting observations in the field to provide a thicker description of the nomads’ interaction with their environment (Flick et al., 2004). One possible avenue for future research would be an ethnographic longitudinal design using digital methods, which is ideal for investigating how the nomadic workers learn to enactment affordances and how they change over time.
Second, our sample was restricted regarding age group as we were not able to find respondents older than 40 (with one exception). This creates a boundary condition (rather than a limitation) for the theoretical
79