The purpose of this paper is to discuss about scientific criteria of tools in qualitative and transformative approaches, specifically activity clinic (clot, 1999). The focus of this discussion is the data processing procedure because of the shortcomings our researchers team often felt at this phase of the studies. The aim is to point out the problems of interpretation that arise and how to take them into account. The argument and the tool that we had built to face this problem are exposed in this article with an example: a recent study about the risk of dropping out in a french low-secondary school and about the way teachers could prevent it. The activity of four teachers has been analysed during one year in their classroom (activity clinic method), precisely the interactions between teacher and pupils at risk of dropping out of school. At first, the theoretical foundations and the conditions of implementation of the intervention research are outlined. The emphasis is specifically placed on the data processing method from the collected linguistic material. The proposed processing implement, broken down into four stages, is to identify precisely in interview transcripts signs and indicators that show teachers’ development and their theoretical underpinnings. It combines all the methodological items founded by researchers in activity clinic for 15 years. The reliability of these items is determined by the precision of their definition and the recourse of a second researcher working independently on data extracts. We illustrate our methodology framework using a case analysis extracted from our research about dropping out. This example of research and the method principles that seem to stand out suggest raise questions on the generalisability of results in clinic methods and on the position of the researcher in the approaches that are supposed to product new knowledge and to help the professionals to improve, to increase their acting power and feel better at work. These questions about the impact of the subjectivity of the researcher on the one hand, and on the other hand generalization from clinical studies are discussed at the end of article.