Page 75 -

60 MADSEN AND KAUTZ
by creating a research data repository (RDR) that would contain details of companies and produc-
tion volumes in the drinks industry. The Web project lasted twenty-four months. The two cases
have been selected for cross-case comparison because of their likeness as “the juxtaposition of
seemingly similar cases by a researcher looking for differences can help break simplistic frames”
(Eisenhardt, 1989, p. 541).
The roles and length of stay in the field have varied for the authors of this chapter. In the Mul-
timedia project, one author was involved in the project as an action researcher throughout the
twenty-two-month time period. This author participated on the development team as the overall
project manager and documented the development process in several ways. Minutes were taken
from all meetings and shared with all involved. In addition to product and process documentation,
data were collected in the form of the researcher’s personal diary as well as statements from e-mail
and informal conversations. Finally, the project contract, the official project progress reports, and
the final project report were available for this study. In the Web project, one author followed the
RDR project closely during the two-year time period. A variety of documents such as the original
project proposal, minutes of quarterly steering committee and monthly project team meetings,
company documents, as well as project reports and deliverables were collected. Furthermore, the
author participated actively in the project as an “action case” (Braa and Vidgen, 1999) or “involved”
researcher (Walsham, 1995) for six months, contributing primarily to the requirements analysis
activity. During these six months, as many details as possible were recorded in the researcher’s
personal diary. In addition, the study draws on seven semistructured interviews with employees
of the case organization, the development team members, and the involved researcher. The in-
terviews were conducted by the other chapter author acting in the role of an “outside observer”
(Walsham, 1995). Individual case study accounts of both the Multimedia and Web cases have
previously been peer-reviewed and published (Kautz, 2004; Madsen, Kautz, and Vidgen, 2006).
These earlier published readings as well as unpublished write-ups have been included as relevant
data material for this chapter, where our purpose, in line with Eisenhardt (1989), is to continue
the work of theorizing from case study research by looking at and beyond the individual studies
to identify patterns across the two cases.
In keeping with the research topic and interpretive approach, our data analysis and understanding
of method emergence has come about through an iterative process of interpretation, comparison,
and interlacing of prior research and empirical data. The framework presented below has been
modified and refined over time according to the lessons learned from its use as a theoretical lens
for understanding emerging change processes in practice (see, e.g., Kautz and Nielsen, 2004;
Madsen, 2004; Madsen, Kautz, and Vidgen, 2005, 2006). For this chapter all data material have
been reread and analyzed anew with an eye to the differences and similarities between the two
cases, and single-case summary and cross-case comparison tables have been outlined. This chapter
presents the findings relevant for understanding how and why the methods emerged differently
in the Multimedia and Web cases.

THE THEORETICAL FRAMEWORK

The ISD literature reveals a lack of cumulative frameworks that integrate the theoretical and
empirical findings from the many existing studies about ISD and ISD methods in practice. No-
table exceptions are: the NIMSAD framework, which is based on both theory and practice and
can be used to select and evaluate primarily formalized methods (Jayaratna, 1994); the social
action model of situated information systems design derived from a case study of a practical de-
sign process (Gasson, 1999); an integrative framework of the information systems development