[Initial approach of an electroencephalographer to statistics].

About concrete examples set by an electroencephalographist (determination of normal values, comparison of data, realization of trial), different methods of uni- or bi-dimensional statistical analysis using qualitative and quantitative variables were presented and the conditions of the tests applications were emphasized. The authors have insisted especially on the necessity of a precise methodology. It has to be elaborated by the statistician together with the physician before the beginning of any trial: statistics cannot be summarized to a simple formula application at the time of final exploitation of data.