Credibility and Validity in Action Research Inquiry

Credibility and validity in conventional social science research function as the researchers’ amulet. In a world of confusing information, these practitioners seem to find comfort in an elaborate methodological (and deeply ritualized) apparatus that purports to resolve these thorny problems cleanly. By focusing on methodological rules as a substitute for facing the question of whether a specific understanding is worth believing enough to act on it, it permits con­ventional social science to bypass the challenge of workability.

It is important to understand what the prerequisites for someone believ­ing in meanings constructed through AR are. We define credibility as the argu­ments and the processes necessary for having someone trust research results. We can distinguish two different types of credible knowledge. First, there is knowledge that has internal credibility to the group generating it. This kind of knowledge is fundamentally important to AR because of the collaborative character of the research process. Its direct consequences in altered patterns of social action constitute a clear test of credibility, a test that many abstract social science frameworks lack. Members of communities or organizations are unlikely to accept as credible the “objective” theories of outsiders if they can­not recognize their connection to the local situation or because local knowl­edge makes it clear that the frameworks are either too abstract or simply wrong for the specific context.

A second kind of credibility involves external judgments. External credi­bility is knowledge capable of convincing someone who did not participate in the inquiry that the results are believable. This is a complex matter. Because AR depends on the conjugation of reflection and action and the cogeneration of new knowledge in specific contexts, conveying effectively the credibility of this knowledge to outsiders is a difficult challenge. Often AR reports are called “mere storytelling,” an insulting attempt to disqualify the general knowledge gained in a specific AR case. Narratives are indeed central to AR. A great deal is at stake in understanding the stories of individual cases in ways that can and should have powerful general effects. Telling stories is not in contradiction to doing social science. It is fundamental to it.

We want to emphasize that the logic of scientific reasoning requires that any individual AR case that contradicts a general social theory thereby invali­dates that theory and requires that a new theory be developed to take account of it. Viable theories do not have exceptions; they must be reformulated to include the exceptions in a coherent way. Thus, individual cases and stories, the stuff of many AR writings, have immense power to alter theories, and theories, no matter how complex or how prestigious a genealogy they have, cannot over­run contradictory cases.

This is the crux of the credibility-validity issue in AR. The conventional social research community believes that credibility is created through general­izing and universalizing propositions of the universal hypothetical, universal disjunctive, and generic types, whereas AR believes that only knowledge gener­ated and tested in practice is credible. Conventional social research believes that only a community of similarly trained professionals is competent to decide issues of credibility, while AR places emphasis on the stakeholders’ will­ingness to accept and act on the collectively arrived at results and the defining characteristic of credibility.

1. WORKABILITY

In AR processes, a first credibility challenge relates to the solution of the AR problem under examination locally. Here the workability test is central. We must figure out whether the actions taken in the AR process result in a solu­tion to the problem. This is in line with Dewey’s ( 1976) thinking on the inquiry process, where knowledge is created or meaning is constructed through acting on the environment. Johannesen (1996) develops a similar conception when he addresses the validity standard of AR. Thus, borrowing from pragmatist thought directly (Diggins, 1994), we understand the inquiry process as an integration of action and reflection and the test of the tangible outcome as workability.

2. MAKING SENSE

The second and complementary process in inquiry is making sense out of these tangible results. How can the outcome be integrated in a meaning con­struction process that creates new knowledge? Here we focus on how meaning is constructed through deliberative processes. Berger and Luckmann (1966) represent an early line in the argument that all knowledge is socially con­structed. Their constructivist position does not reflect sufficiently on the qual­ity of the socially constructed knowledge, however, because they do not attempt to scrutinize the quality of the constructed outcomes. For them, any construction is as right or wrong as any other possible one, a position anti­thetical to AR. ‘ In AR, we use processes in which chains of arguments can undergo some kind of testing procedure. We can describe two possible processes for such deliberations: Habermas’s (1984) ideal speech situation and Gadamer’s ( 1982) hermeneutics, though many more formulations exist.

The Habermasian ( 1984) ideal speech situation counterfactually character­izes a process free from domination, where the actors involved in meaning con­struction exchange arguments without coercion. In this idealized situation, each participant seriously and honestly judges the arguments presented to him or her and comes back with the best judgment he or she can make in the response argu­ment. This process leads to an understanding that is characterized as a legitimate truth when no further arguments are able to overturn those already stated.

The credibility of this line of argument emerges out of this ideal situation when no better explanation can be offered. This is not a one-shot affair but more a continuous process in which new experiences or new arguments con­tinually challenge what is already thought of as credible knowledge. Scientific knowledge is then easily understood as an ongoing discourse searching for the best interpretation of certain phenomena. The Habermasian (1984) ideal speech situation is alluring in its strict logical and rational reasoning, but it leaves out emotions, power, and inequality as key determinants of all commu­nication processes, a critique that has been widely made (see, for example, Flood and Romm, 1996).

Gadamer’s (1982) major work, Truth and Method, resists facile synthesis. He treats Habermas’s ( 1984) ideal speech situation as a piece of naive idealism, advocating instead a more complex combination of dialogue, mutual interpre­tation, and eventual (but never final) “fusion of horizons.” Gadamer respects the historicity of the knowledge, interpretations, and experiences the partici­pants bring. Gadamer, unlike those who have tried to convert hermeneutics into an academic parlor game, insists that hermeneutics is a form of acting, not merely a method for thinking.

3. TRANSCONTEXTUAL CREDIBILITY

At a still broader level, there exists the possibility of transcontextual mod­eling of situations, and this can be explained historically and causally. This is vitally important because it is precisely here that conventional social scientists usually invoke the canon of generalizability and try to move social research toward what they view as objectivity and away from what we understand as scientific research.

Our view, paralleling that of Francois Jacob ( 1982) on the possible and the actual, sees every situation containing more possibilities than those that are acted on. We understand that all current situations could have been different but were not. A particular outcome was realized through the intersection of environmental conditions, a group of people, and a variety of historical events, including the actions of the participants.²

From this perspective, all explanations of present situations are actually accounts of historical moments and particular causes acting on particular organizations in specific contexts. In this way of thinking, theory does not pre­dict the outcomes of a particular situation. The role of theories is to explain how what happened was possible and took place, to lay out possible scenarios for the future, and to give good reasons for the ones that seem to be the prob­able next outcomes. This latter move, of course, relies precisely on analogizing outcomes from other cases and contexts in a coherent way.

These practices are science at its best, and our previous example was drawn from evolutionary biology. Are there examples in the social sciences? We believe there are, but that they are generally ignored. One such example is the structure of thinking underlying the work ofMax Weber. He built a wide variety of ideal types to deal with the diversity and complexity of the issues he studied: bureaucracy, charisma, legitimacy, authority, religion, urbanization.- In every case, he created an abstracted list of transcontextual characteristics after the painstaking study of many historical cases. He then used these char­acteristics to develop explanatory strategies.

Weber’s (1958) rarely cited work on cities presents a particularly clear example. He gathered all the evidence he could from all over the world about the phenomenon of cities. On the basis of this broad reading, he developed a synthesis of the traits he found in each major example of cities in different places. He then took this list of traits and arrayed the traits together until he had a list of all the major features that he could find in the cities of the world. The total list of major traits made up the basis of his ideal type of city.

This was only the beginning, however. Armed with this list, Weber returned to each world area, to each context, to examine what traits were pre­sent or absent in each situation. When he found particular complexes of traits present or absent in a location, he reexamined the history of that area to explain the presence or the absence. Gradually, he developed what he calls a “causal interpretation of history” that helped him understand why particular features were present or absent in particular situations, built over the backdrop of a general knowledge of the phenomenon of urbanization.

Weber’s way of examining particular situations and environments by closely gathering traits from those situations, listing and analyzing those traits, and then returning to the particular situations helped him understand why particular features were present or absent. This is how knowledge developed in one AR situation is to be transferred to other situations. AR does not general­ize through abstraction and the loss of history and context. Meanings created in one context are examined for their credibility in another situation through a conscious reflection on similarities and differences between contextual fea­tures and historical factors. They are moved from the context where the under­standing was created through a collaborative analysis of the situation where this knowledge might be applied. Based on the historical and contextual analy­sis, AR judgments are made about the possibility of applying knowledge from one situation in another. This is also the proper way to develop AR approaches, through the in-depth analysis of the use of a variety of techniques and processes in multiple contexts.

Thus, we believe that detailed attention to cases, context, and history is essential to the development of science in general. It is the most meaningful way to proceed in developing a social science that respects the diversity of sit­uations while also developing an understanding of the processes found in many situations and that can be used to explain what happened in each case.

Source: Greenwood Davydd J., Levin Morten (2006), Introduction to Action Research: Social Research for Social Change, SAGE Publications, Inc; 2nd edition.