Teaching Iterative Design in Action Research

An instructive example from Bangladesh in the Action Research Journal

​Last year, I was invited by the editors of the Action Research Journal (ARJ) to author one in a series of short papers which offered reflection and feedback on the articles published in certain volumes of the Journal. The May 2017 edition of ARJ features this format, in which the feature articles are followed by these brief response papers.
 
The article I was asked to review was written by five authors from Vrije University in Amsterdam: Anastasia A Seferiadis, Sarah Cummings, Jeroen Maas, Joske GF Bunders, Marjolein BM Zweekhorst. Their article, “From Having the Will to Knowing the Way: Incremental transformation for poverty alleviation among rural women in Bangladesh” focuses on a women’s livelihood training program in northern Bangladesh. From an action research practitioner perspective, their article focuses on the iterative, adaptive design of this program over a span of more than a decade.
 
My response paper, available at this link, noted a few limitations with the group’s approach, namely that the team’s process was action-oriented and adaptive but not participatory. The progressive changes in the program made over years were not driven by direct feedback from the participants but rather by observations of the authors/program directors. Observational data and participatory data don’t exist in mutually exclusive universes; certainly there would have been much overlap. However I felt the program could have been more attune to the participants’ needs and could have evolved more rapidly if the authors had engaged the participants directly by providing roles/mechanisms for them to actively and consciously shape the program.
 
Nonetheless, I highly recommend spending some time with this paper because of the way that the authors have documented and explained the evolutionary development of their program. While I may have some disagreement about the level of participant engagement within the program, I find the narrative explanation of how and why the program was adapted very effectively articulated. As such, I find the paper a useful resource for teaching about action research.
 
Action research, in a world where most people’s exposure to research approaches is limited to basic positivism, doesn’t seem to them much like research. In this narrow view of research, circumscribed by the concept of the scientific method, rigor is defined by setting up a strict, clear series of steps which are followed without deviation. While this variety of research is important and valuable, it is certainly not the only way to conduct research, particularly when working with people in an effort to shift their social dynamics and quality of life. Nonetheless I find much resistance amongst students and colleagues when I suggest that they can change their action research process in mid-stream if it isn’t meeting their aims. The mantra “if it’s not working, change it” seems too good to be true. Indeed good research to them is exemplified by a lemming over a cliff mentality; even if it’s going to fail you must follow through and document the failure. From this perspective, no wonder that most people see research as rarified, in that you have to know everything and anticipate everything before you even start. Certainly there are disciplines where such failure is ok and can provide important learning; however, when dealing with people’s lives and livelihoods it’s far better, and far more ethical, to make changes as soon as the better path becomes apparent.
 
For this reason, I find this paper from the Vrije researchers quite instructive and informative as a teaching aid, for demonstrating the process of iterative research design in action research. I have taught iterative research design several times this past year to USIP Generation Change colleagues in Kenya. The idea of iterative design, even when explained and justified, still seems counterintuitive to most people’s native sense of how research is supposed to work. The Bangladesh paper goes a long way in helping to shift that perspective. While the paper contains several dimensions and lines of analysis, almost of half of the content is focused on adaptive design. There are 3 charts that convey the design components in different ways, by timeline, by learning cycles, by the key activities within each cycle, as well as through a highly useful graphic which combines all of these elements at once.

One of several explanatory graphics from Seferiadis et al's article in the May 2017 edition of ARJ.

​Moreover, in the text of the article, the authors provide extensive detail of the situations within each learning cycle that drove specific changes at each key moment. In Empyrean’s action research trainings, we walk our students through each of these different phases from the article and so that they clearly see the linkages between the key situational/programmatic/contextual issue and the change in the methodology that resulted. That the project is a relatively straightforward livelihood project also helps—one location, one group of participants over time with a finite, clearly understood goal.
 
Earlier in the Empyrean training module we will have explained the concept of recoverability as a key element of quality and integrity in action research. It’s initially a challenging term for students/team members to grasp, but the Vrije paper, with its well-differentiated learning cycles, graphics and timelines, provides a concrete example of how recoverability can be evidenced thoroughly and convincingly.
 
I commend the writers for their efforts to articulate and visualize their iterative design process so thoroughly. For those who teach action research approaches in academic or project settings, I recommend this paper as a valuable addition to your reading list and as an instructive example to deconstruct thoroughly in class.
 
What other examples/case studies do you use to help your students and/or co-researchers better understand the iterative design process?