The Causality has been always an intriguing question and there have been always debates on how to deal with it. I will introduce basic concepts based on the counterfactual approach and present a simple operational tool to clearly define graphically a causal model. Often, particularly in medical and epidemiological studies, even now, it is entirely clear if the data at hand allow to answer the research question of interests and it is not unusual to read papers where even if an etiological setting the statistical approach is based on semi-automized procedures. This talk introduces the importance of Directed Acyclic Graphs (DAGs), which induces the scientist to formulate with arrows and lines
the dependencies among the variables of interest, with the ultimate goal of determining if we can aim to possibly draw causal conclusion from the study. In fact, DAGs represent a powerful tool not only in the statistical analysis of the data, but often shall be considered in the design phase, not being required in many grant applications, such as NIH in the US. DAGs were developed around 20 years ago by Judea Pearl, from Stanford. In fact, his book on Causality is considered a milestone in this field and his webpage is a rich source on information for new learners. Behind the graphical approach there is a formal and elegant mathematical theory, based on structural equation models, which I will introduce if time allows.
9 Ottobre 2026, ore 12
Rino Bellocco, ScD
University of Milano-Bicocca and Karolinska Institutet
n person: Room 24 (4th floor) building CU002 Scienze Statistiche
Webinar: https://uniroma1.zoom.us/j/83625004899?pwd=bXCtz0mp759PUh2lkqT0BUoVa0Uegg.1
ID riunione: 836 2500 4899
Passcode: 123456