Apports et limites de l'approche statistique

The interest for quantitative approach in research is increasingly growing especially with the development and availability of software. This increasingly systematic usage questions us not in terms of relevance but in terms of validity of the results. Can we be sure that the data from which decisions are made truly reflect the greatness and meanings they imply or are they purely formal constructions with no real base? The problematic questions the epistemological and technical aspect by the use of superficial analysis of different levels of the statistical approach (descriptive, inferential statistics and modeling). A special attention has been paid to remove from this article any mathematical reasoning in order to avoid the maze of mathematical consideration. The article is not looking for a conclusive answer but put the emphasis on the issue and debate while hoping to highlight the criteria for a successful quantitative research and safe use.