Statistics has been traditionally a difficult subject to teach and learn. Even the students with good mathematical skills may have problems in understanding the main principles of randomness, distribution functions and statistical significance. Additionally, it is not strange that students which have a good understanding about probability theory does not know how to use it in the statistical problem (i.e. once a random sample has been obtained, and the aim is to infer the values of parameters or distribution functions which better model the obtained data, in order to make decisions based in this information). This paper proposes a work in progress to improve the student understanding and involving in the process of statistical decision making, participating in a statistical decision making laboratory. The method is based in the following principles: The use of real economic data in order to predict the behavior of economic indexes, allowing the students to compare the real results with the results of the laboratory work. This comparison should be realized in a formal way, using decision theory principles (hypothesis test methodology). The use of the same statistical tools learned in the course in order to analyze and compare the results presented by the students. That is to say, if the numeric results from different students or student groups are very sparse, this fact perhaps indicates a por work definition by the teacher. If the results obtained are very homogeneous but are far from the real world results, this perhaps indicates a common basic problem in the presentation of the theory and the laboratory work to the students. This is an important point, since the students use the tools used in the laboratory to evaluate the performance of their laboratory results, and to propose improvements, which is a clear motivation point.
