Extracting case study research and multiple contexts?

The following is an attempt to consider and respond to a suggestion on my thesis from my esteemed supervisor. It’s essentially some thinking and questions arising from further refinement of the research method being used in my thesis.

The suggestion was that perhaps my work is an example of van Aken’s (2004) extracting/ive case study research. Is this the case?

What’s the method?

I’m essentially using an action research method within a single case study to develop/abstract a design theory for e-learning. The action research cycle is somewhat disjointed in that the case – the development and evolution of Webfuse from 1997 through 2009 – involved an on-going process of action research as different aspects of the system were developed and changed (evidence of this is found in my publications where this process was played out).

In addition to this more traditional example of action research there was another round of action research which occurred after most of the initial Webfuse development was completed. This was the cycle that led to the development and re-development of the design theories that were abstracted from the Webfuse. This is almost an example of a meta action research cycle. The existence of this on-going cycle is also evident in the publications from about 2002 onwards as there are three different representations of the ISDT, with another being formulated now.

Extracting and developing case study research

This type of method does seem to be fairly evident in the broader business research methods with van Aken (2004) identifying two relevant types of case study:

  • extracting case study research; and
    Also labeled as “best-practice” research. The aim here is to extract technological rules (theory) that is already being used in practice.
  • developing case study research.
    Researchers collaborate with people within specific contexts to solve a local problem. However, there is a reflective cycle after each case in which the researcher seeks to develop knowledge that can be transferred to similar contexts. Also labeled as clinical research.

It appears that van Aken (2004) makes another distinction between these two in that extracting case study research generally involves multiple case studies in which the same best practices have been implemented in multiple contexts.

However, van Aken (2004) does make the suggestion that developing case study research isn’t necessarily multiple case study. Instead development can go through an “alpha test” in which the “rule” is evaluated within a single context. However, significant value arises when there is “beta testing” when the “rule” is translated into other contexts and “third parties use it, assess its effectiveness and make final improvements”. van Aken (2004) makes the connection with beta-testing and replication research.

However, this bit is interesting

An essential element of beta-testing is that testing is conducted by a third party to counteract the ‘unrecognized defences’ of the originator of the rule, which may blind him or her to possible flaws in its use (Argyris, 1996).

This somewhat later quote is also of interest to my thesis

Management Theory research as described here may have most in common with the approach to Action Research discussed by Eden and Huxham (1996), who stress among other things the need for generality of research results (p. 530) and write about ‘action research aimed at the study of organization and organizations . . . where it is likely that the researcher accepts the dominant managerial ideology’ (p. 529).

So which “case study research” is my thesis?

From this description, I think that my thesis research leans more towards the developing case study research. There has, however, been two different reflective cycles. The first was the immediate reflection on the intervention that contributed to changes in the intervention. Then there was some additional reflection afterwards to move into the design theory. It was this additional reflection that is aiming to generate the more general contribution to knowledge.

So when does a context change?

An interesting aside here is exactly how do you define a different context. To some extent it, I would argue that CQU in 1996/1997 – when Webfuse started development – is almost an entirely different context in 2001/2002 and certainly by 2005. It’s still the same institution, however, the internal structure, people and processes have changed quite significantly. Significantly enough to be a new context? Not sure. But answering that question might be interesting.


van Aken, J. (2004). “Management research based on the paradigm of the design sciences: The quest for field-tested and grounded technological rules.” Journal of Management Studies 41(2): 219-246.

PhD Update #28 – minimal work, feedback arrived

Am attempting to keep this weekly update thing going within the new context. Not much to report this week as I was out of town for three days, baby sitting and recovering for another, and thinking about jobs and family-based travel on the other.

Am going to try a more evolutionary, to do list oriented approach to this post as well. i.e. the finer grain tasks I need to complete will be added and hopefully checked off here during the week.

What I did last week

Two fairly small bits of work done this week:

  • Completed a draft of the “lessons learned” section of chapter 5.
    This is still fairly early and will be revised, but wanted to get something down on paper, so to speak.
  • Received and started processing some feedback from my esteemed supervisor.

This left at least 3 of the tasks I set myself, undone.

What I need to do next week

A growing list of tasks to do:

  • Read comments on chapter 6, modify what I’ve done, send it off to supervisor.
    • Main point is to have a good overview of findings, i.e. summarise the Ps Framework, New view on ISDT and the ISDT for e-learning.
    • Look at including the Gregor and Jones table for theory summary.
    • Implications for IS/Design researchers needs to mention contribution of Gregor and Jones
    • See what insights Eisenhardt (1989) can add.
    • Look into van Aken’s “extracting” and “developing” case study concepts and see how it can be worked into this chapter and also perhaps chapter 3
    • The conclusions section needs to be more of a “declaration of victory”.
  • Read comments on chapter 4, especially on the ISDT
    • Think about how chapter 4 ISDT may change in response;
      • Table 4.3 – the principles read more like requirements. This opens up the question of what form should a principle take? Perhaps Bunge/van Aken’s technological rules?
        van Aken (2004) “If you want to achieve Y in situation Z, then something like action X will help”
      • Is section 4.3 about high-level guiding requirements and Section 4.4 more about design principles? YES
      • Need to think about the names of the principles/guidelines, especially the important/innovative ones.
      • Add the Gregor and Jones “theory table” as a summary/overview
    • Start thinking about form of ISDT for chapter 5.
    • Make more general changes to chapter 4 and the rest of the thesis
      • Rework the Chapter 4 introduction, and perhaps other parts of the chapter, to better orient the reader to the purpose and the structure of the chapter.
      • Look at how much of the design work in the thesis was done prior to commencing the thesis and how this is outlined within the thesis, with some connection to the extractive case study conceptNO.
  • Complete evaluation section of chapter 5.
    • What can be evidenced about the trend in feature and broader user adoption about Webfuse systems post 2004/2005.
  • Revisit chapter 1 and get to final draft, start sending around.