Data-driven research

Data-driven action research



This is a resource file which supports the regular public program "areol" (action research and evaluation on line) offered twice a year beginning in mid-February and mid-July. Orientation material for the course is typically emailed two weeks prior to the start.

For details email Bob Dick bd@bigpond.net.au

... in which I describe a process which allows both the process and the content of action research to be driven by the data in a style somewhat reminiscent of grounded theory

A French translation by Anna Chekovsky is available here.
A German translation by Daniel Gruber is available here

You can think of action research as a double helix. (That’s where the logo used for these pages originated.) As the research proceeds, two intertwined spirals develop, eventually to yield ...

a better understanding of the situation being researched, and as a consequence better plans for action and change;

better processes for researching the particular situation: more penetrating questions, processes which better suit the local culture, and so on. ¹

After a brief discussion of each of these, it is mainly the first that I address in this paper.

As I have said elsewhere, at the beginning of a change program it is hard to know enough to design the program in detail. Wherever you start is the wrong place. Why not postpone the detailed design until you better understand the situation? This can lead to better designs, and in turn better data and understanding.

Research and change might seem to be a difficult combination. In practice, however, each can inform and enhance the other. As with the other double helix, the two helices are not only intertwined. They are also interconnected. Better understanding allows better designs, leading to better data and in turn yet better understanding.

In a style similar to grounded theory research ² your growing understanding can develop in such a way that it is driven by the data that you have collected. This can provide some protection against the biases and preconceptions which researchers and others inevitably bring to the research situation.

At the heart of the action research processes I use is a simple strategy. When there are two or more pieces or sets of data there are likely to be disagreements between them. The data sets may arise from different case studies, different data collection methods, different questions which pursue similar information, different informants, the same informants at different times, different theories guiding the data interpretation, and so on.

Properly understood, most of these differences turn out to be illusory. Most of them are artifacts of the different situations or methods or people. The data-driven process we are exploring here therefore seeks to bring about a different understanding. It does this by noting the differences and seeking explanations for them.

In short, the apparent disagreements develop the understanding.

The process can be strengthened by increasing the number of disagreements. To do this, note agreements and then seek exceptions. Those exceptions become the disagreements that fuel the drive towards a deeper understanding. Agreement becomes disagreement becomes a deeper agreement.

In more detail ...

For ease of explanation, consider two data sets perhaps arising from two different informants. The information then fall into three categories:

mentioned by only one of the informants;

mentioned by both informants, and in agreement; for instance both of the informants mention the diversity in the organisation or community, and both with approval;

mentioned by both informants, and in disagreement; for instance both mention diversity, one approvingly and one disapprovingly.

Here is what you do ...

For the moment, ignore the idiosyncratic information unless there is some compelling reason not to do so. If the information is important it will emerge later. It is unlikely to be lost.

Note each apparent agreement. Thereafter, for each, look for exceptions.

Note each apparent disagreement. Thereafter, look for explanations for each.

I expect that you noticed that an exception disagrees with a former agreement. It turns an apparent agreement into a disagreement. An explanation resolves an apparent disagreement, turning it into a deeper agreement. Another double helix.

Here is the process diagrammatically:

I might then add that, after action, the results of that action are reviewed. This provides more data, beginning the cycle once again.

Notes

I could also have chosen other pairs of helices, most importantly "theory and practice"; and, as will be seen later, "agreements and disagreements". [ back ]

The seminal work on grounded theory is Glaser, Barney G., and Strauss, Anselm L. (1967) The discovery of grounded theory: strategies for qualitative research. Chicago.: Aldine. [ back ]

_____

Copyright (c) Bob Dick 2000. This document may be copied if it is not included in documents sold at a profit, and this and the following notice are included.

This document can be cited as follows:

Dick, B. (2000) Data-driven action research [On line]. Available at
http://www.aral.com.au/resources/datadriv.html



Maintained by Bob Dick; this version 1.02w last revised 20140510