Causal inference based on counterfactuals--Detailed interpretation of 16,000-word literature (reasoning application of causal relationship) [full text summary]

Foreword:

        During the summer vacation of Research 0, this article is also an ending to part of my study in the past two months!! During this period of life, I have experienced discomfort, experienced confusion and could not find a learning method that belongs to me. . Writing down the interpretation of this article also made a summary of my recent time, and I hope that I can persist in my postgraduate life in the future! Maintain the exacting standards you hold yourself to now! ! Keep yourself unyielding, not reconciled! ! I also hope that this article can always motivate myself --- "There is no beginning, but there is an end!"

Article frame:

1. Research Background and Significance:
 

①In epidemiological and medical research, counterfactual or potential outcome models have increasingly become the standard for causal inference.

② Counterfactuals are the basis of causal inference in medicine and epidemiology.

③ Difficulty: In observational studies, it is difficult to estimate counterfactual differences.

④ The only necessary condition for a causal effect on an individual is the priority of the factor's influence on it.

⑤100% evidence of causality is impossible.

⑥ Question: How much evidence of causal effects can one gather in practice, and what statistical models can contribute to this evidence.

⑦Author's opinion: The counterfactual model of causal effects captures most aspects of causality in the health sciences.

2. Use in the article