The focus of the course will be on interpreting the links between variables from different data sources, taking into account the theories set out in the other courses. The course will begin with a presentation of different existing data structures: cross-sectional data, time series, or longitudinal data. The course will then primarily focus on the classical linear model and its statistical properties, including the choice of a functional form, hypothesis testing, the use of dichotomous variables and the interpretation of estimation results. Various issues must be addressed, including the problems of serial correlation, heteroscedaticity, and endogeneity. The advantages and disadvantages of each of the proposed solutions will be discussed. Each facet of the course will be practically applied using real data.