A Second Chance to Get Causal Inference Right: A Classification of Data Science Tasks
For much of the recent history of science, learning from data was the academic realm of statistics,1,2 but in the early 20th century, the founders of modern statistics made a momentous decision about what could and could not be learned from data: They proclaimed that statistics could be applied to make causal inferences when using data from randomized experiments, but not when using nonexperimental (observational) data.3,4,5 This decision classified an entire class of scientific questions in the health and social sciences as not amenable to formal quantitative inference.
Some content is only viewable by ASA Members. Please login or become an ASA member to gain access.