In line with the majority of studies employing a reconfiguration or ANT lens, multiple methods for data collection were used to capture the complexity of the phenomenon.
Initially, the analysis of documents – including, for example, project documentation, organization charts, role and process descriptions, and intranet pages – provided the
ba-sis for identifying relevant topics and participants for interviews and observations.
These documents also provided access the company internal concepts and terminology early in the study and thus facilitated the discussion with participants especially in the initial phase. Participant observation (Gold 1958; Myers 2008) of clerks at CF aim at the development of an understanding of the work environment and the context within which participants interacted with (O)LMS and each other. That is, it helped capturing the situated material and social nature of the phenomenon and thus relates to the per-formative aspect. But, it also involved asking questions as well as brief, informal dis-cussions with participants, which enabled gaining access to their perceptions and under-standing of these situations. In addition, the more formal interviews provided insight into the understanding of the phenomena of interest in general (i.e., the ostensive as-pects). Still, interviews were not limited to abstract ideas, but also allowed to learn more about the personal experiences and thoughts of the different individuals studied (Schultze and Avital 2011). This was facilitated by the semi-structured design of inter-views (Myers and Newman 2007) that was chosen to “generate deeply contextual, nu-anced and authentic accounts of participants' outer and inner worlds […] their experi-ences and how they interpret them” (Schultze and Avital 2011). In order to develop a comprehensive understanding of the phenomenon and to avoid “elite bias” (Myers and Newman 2007), individuals from all levels of the organizational hierarchy were inter-viewed.
Two researchers were on-site at BANK about one year before go-live of LMS exploring the overall ES implementation program. The exploration primarily involved analyzing project-related documents and selectively talking to members of the program team. The primary data collection was done in three waves starting two month before LMS go-live. Every wave involved at least one week-long visits to CF. Figure 15 shows the tim-ing of the waves relative to the go-live of LMS.
Data collection
Figure 15: Timeline of research activities
Wave one was conducted to capture the original routines at CF before LMS was intro-duced, that is, the status quo before go-live. In November one researcher participated in a one-day basis training to gain some knowledge about the technology to be introduced at CF and experience the preparation CF employees had before go-live. In December two researchers performed 21 interviews as well as participant observations to capture the way people worked at CF before LMS was implemented. All participants at the clerk level as well as a team lead and one member of top management were also ob-served during their work after each interview for at least one hour resulting in 17 partic-ipant observations.
Wave two captured the immediate time after go-live, as well as the subsequent two months. One week after go-live in January (Wave 2a) the same two researchers con-ducted observations of respondents of wave one and the situation at CF in general in-cluding participation in telephone conferences and status meetings. Interviews with clerks were not scheduled during this time because CF employees were very busy with the effects of the ES technology introduction. Still, two interviews with a team lead and a top manager were conducted during the visit. The second complete round of 19 inter-views was conducted in February (Wave 2b), about two months after LMS was intro-duced.
Finally, wave three was conducted in July to capture the situation six month after the system was introduced and employees have had sufficient time to accommodate the new technology. Overall, 15 interviews were performed in this wave and most employees whose work was already attended in waves one and two could be observed while work-ing.
Participant observations typically took about one hour, during which a single individual was observed while working. Still, because most participants were working in
open-plan offices, observation was not limited to the individual currently accompanied, but interactions between other CF employees frequently could be observed as well. In total 38 observation sessions were conducted. In addition, several participant observations (especially in wave 2a) took several hours. For example, a group of CF employees were accompanied for one whole day in January, who were selected to support PSN and thus were required to learn how to use LMS functionality relevant for their new tasks.
Observations were captured directly during and after participant observation as well as at the end of each day on-site similar to Volkoff and colleagues (2005) and were fre-quently compared and discussed by the two researchers present at CF. Additionally, de-briefings with a colleague, who was not on-site, were conducted to discuss and docu-ment experiences and observations.
The interviews were conducted in German and took between 30 and 120 minutes. All interviews were tape recorded and transcribed. For the interviews two types of interview guidelines were developed, one for managers and one for clerks that were supposed to work with LMS directly. During wave one, interview questions focused on participant’s work processes and routines, their interactions with colleagues, the role of technology in their work, and expectations towards LMS and its impacts. In wave two interviews cen-tered on changes in their work practices and routines, in the technology supporting their work, and the problems that occurred after LMS go-live as well as the actual impact the implementation had form them. Similarly, wave three was intended to capture the ef-fects of the implementation with regard to participant’s way of working, social interac-tion, and technological support, but further focused on the emergence of the solutions established by then. Each interview in the three waves closed with an open question asking participants to freely talk about any themes they deemed important for under-standing the situation at CF as well as their experiences during the change in general.
Table 8 shows the distribution of interviews over the three waves and the departments at CF.
Overall, the case material comprises 1185 pages of interview transcripts, 653 pages of hand-written observation memos and field notes as well as numerous documents gath-ered during the field visits. All material as well as intermediary results of the analysis were stored in a research database (AtlasTi, Microsoft Sharepoint, Excel, and Visio) from which the results presented in the subsequent chapter were derived.
#Interviews
Department/position # Persons Wave 1 Wave 2a+b
Wave 3 Total
Top management 4 3 5 2 10
SSG team lead 2 1 1 1 3
SSG clerks 8 8 7 4 19
PSN team lead 1 1 1 1 3
PSN clerks 4 4 3 3 10
PSS team lead 1 1 1 1 3
PSS clerks 3 3 3 3 9
Totals 23 21 21 15 57
Table 8: Interviews at CF