Authors: Alfredo Berbegal, Researcher, Department of Methods of Research and Diagnostic in Education (MIDE), University of Zaragoza, Spain
Fernando Sabirón, Research Director, Department of Methods of Research and Diagnostic in Education (MIDE), University of Zaragoza, Spain
Patrick Boumard, Research Director, Department of Philosophy, Breton and Celtic Research Center (CRBC-CNRS), University of Western Brittany, France

3. Working With Data

The most important challenge in this phase was how not to make discontinuous what is continuous. To be very formalist and define our categories and linkings without doubts and paradoxes only deadens our obsessions and anxieties. Social, personal and professional situations are dynamical, not static, not fix. Inside the own life of fieldwork, working with Data was a rizomatic (Deleuze & Guattari, 1977), non-linear and anarchic (Feyerabend, 1975) process. This revealed the natural incompleteness of our research subject, its complexity.

3.1. Data Divisions
Any "definition of situation" was forever fed by my precedent Observation-Participation game. Such Data divisions was made from my research texts, afterwards thoroughly organized:

• First Division. My Three Texts. I managed three kind of diaries: my research diary, my personal diary (Lourau, 1994) and my training diary. All of them were separately analyzed.


• Second Division. Institutional Texts. I also included other texts like legal documents, section diaries, implemented social and psychological programs, evaluative annual reports... I had to know its effects and its constrictions according to multiple events, situations and practices.


• Third Division. Multiple subtexts. Inside the research diary, I made a lot of subdivisions. These were not categories, but subtexts to be analyzed later. Firstly, my partitions were inspired by one of the most visible rationality, the instrumental one (Weber, 1944): institutional working; activities, times and spaces; sections descriptions; security measures; institutional controls; sanctions and disciplinary measures; minors' routines; problems, conflicts and controversies; institutional tensions... roles and relationships between them.


• Fourth Division. Residual subtexts. Precedent subdivisions were very simple and elemental and I had to polish my analysis. I added three divisions called "Reflexions" where I included: professional representations, minor perceptions, companionship perceptions, existential reasons for institutional intervention and ruptures in the social theory of roles. The principal aim was to grasp the imaginary of this professional context, but exploring other kind of rationalities. I kept precedent divisions, but I was modifying their boundaries.

Precedents divisions and subdivisions were not absolutely comprehensives. Those only intended to respond to a technical question: the analysis. Those partitions were a strategy to start to think. Only that. However, my research was always threatened: the analysis itself could kill the life I just wanted to understand. Divisions could impose a dead reason (Lourau, 1988; 1997; Sartre, 1967). In this way, I was my own hangman. I had to avoid being seduced by the certitude and to be extremely sensitive to the incertitude, variation, to the different points of view, "listening" what the fieldwork tried to say to me.
I constantly changed relations and linkings (Fielding & Fielding, 1986). Categories and analysis units were defined across all the texts and all kind of artificial division. This allowed me to naturalize my analysis, to criticize my absolutist reason and, very important, to keep open my mind to the contradiction of what I eventually was conceptualizing.
This process was always split between the fieldwork - where the significance was in the heat of the moment - and my "cosmovision", scientifically speaking and in a completely nonmystical sense - where signification desperately sought its referent (Kauffman, 1995; Kluger, 2008; Morin, 1986, 1990). Obviously, if this "cosmovision" tried to stretch the notion of person (Rogers & Stevens, 1967) and its social and psychological analytical rationalities, working with Data got really more unstable.

Just a few words about the use of some qualitative software for working with Data. All these texts were imported into NVIVO software (Richards, 2002). Nevertheless, I have to recognize that I used this software as an interactive big folder or a filing cabinet. I ordered texts, subtexts, divisions and sub-divisions. I moved and recovered them according to my new and "floating" criteria. But I could not say that I managed my analysis by this software.

Back to Project Home Page