Two distinct categories can be established in the field of exploratory studies. First, there are studies on subjects that are relevant but which, for various reasons, have been left more or less unexplored. The second category of exploratory studies comes above the first category since these studies revolve around issues which relevance has not been yet supported in a way that would justify thorough consideration and adéquate investment, even if a certain form of argumentation could be elaborated on their potential interest. To work with the second category of exploratory studies, it is necessary to create an approach that will confïrm or invalidate their potential interest and, in the case of a confirmation, determine the orientation of the subsequent studies.

A problem may arise from the use of the last mentioned category of stratégies. When these strategies do confirm the potential interest of the studies, they have to be given up for other strategies that are best suited to carry out the research. Consequently, new resources as well as additional time are required. In both scientifically and professionally oriented fields, this problem could have a negative impact since the researches carried out in these fields often give rise to more immediate concerns and expectations. In our opinion, the problem arising from this discontinuity is not necessarily unavoidable.

In this article, our objective is to set out and examine an investigation strategy that we already used to tackle a problem falling in the second category of exploratory studies and relating to the implementation of technological changes in unionized environments.

This strategy uses concepts such as «concrete action System», «power», and «uncertainty zone» which form the basis of the theory of organizations elaborated by Crozier and Friedberg (1977). This approach avoids the discontinuity problem previously mentioned and indicates interesting investigation paths, when such paths exist. Moreover, it does not require a large investment at the beginning of the research. Finally, it seems to us that this strategy can easily be used and adapted for several research problems that arise in the field of relations in the workplace.