CMStatistics 2023: Start Registration
View Submission - CMStatistics
B0914
Title: No causation without manipulation: The public responsibility of science and policy making Authors:  Marco Di Gregorio - University of Turin (Italy) [presenting]
Zenia Tea Simonella - University of Milan Bicocca (Italy)
Abstract: The concept of cause has ancient origins, going back to Aristotle, but Hume was the first who emphasized its psychological genesis. Other social scientists sustained this idea, criticizing the possibility of experiencing a causal relationship. Apart from the nomological-deductive approach, which introduces the syllogism to solve the issue of causality, other approaches focus on a mechanistic type of causal explanation. In particular, in observational studies and in experimental ones the main issue is the degree of control an experimenter has over the phenomena under investigation according to the Fisher-Rubin-Holland model. In recent decades, causal inference models and experiments have been used in social sciences, especially in the field of public policy evaluation. However, they are criticized both for the lack of respect for all the criteria requested for the experiment (e.g. randomization) and for the assumptions they take into consideration (e.g. the idea of uniformity of the subjects). After having outlined how the concept of causality was conceived, the main critical issues are discussed when using the experiment in social sciences and in policy making. Finally, methodological rigor is emphasized not to be only a technical matter but also an issue of public responsibility in interpreting and using scientific results.