A questionnaire is drawn up according to the issue and the themes raised. From this, a representative sample of the population is selected taking into account the constraints (including precision criteria).
First, we propose a very simple exploratory analysis of the data to get an idea of their quality. The second step will be a bi-varied analysis (taking into account the client’s requirements). And finally, we propose a multivariate analysis ranging from linear regression to analysis of the main components.
Intended to collect qualitative elements, which will not be directly quantifiable by the targeted individuals.
It is most often carried out by focus groups, individual interviews or observations. The objective is to gain an in-depth understanding of the behaviour of the study population.
Multidimensional analysis is one of the most common methods in decision making. These are methods that allow any number of quantitative or qualitative variables to be processed simultaneously.
These methods generally aim to develop forecasting models, segment or discover relevant associations. It is the basis of several data mining techniques.