Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction

Ok

I bought this book for assess experiments in social science field. It is well written and contains all the information required in order to have a good preparation in causal inference. I appreciated the logical path and his linear structure!

Some people are of the opinion that statistics should replace calculus in high school/college curricula because of the importance of analytics in the modern business world. I think a similar argument could be made about replacing some of the more esoteric parts of graduate statistics/econometrics with causal inference theory. If that opinion is correct, this would be an excellent reference text. Causal inference theory is important because the regression techniques now taught to young social scientists as methods of determining cause and effect assume endogeneity when the data often don't support such an assumption. They also impose a linear model on the data that can be similarly inappropriate. The non-parametric techniques discussed by Rubin and Imbens, while having their own assumptions, are applicable to a wider range of problems. Rubin and Imbens summarize the voluminous literature on propensity score and related causal inference techniques in a manner that is accessible to someone with a solid background in statistics (both frequentist and Bayesian). I read the book cover to cover and, despite already knowing something about Propensity Score techniques, learned a great deal. They begin with randomized experiments then explain how the mathematical models developed for such methods are also applicable to observational studies. They then discuss various methods of using the Propensity Score along with tests of the plausibility of such models and bias limits when some of the assumptions in these models are relaxed. One complaint I have is that the different types of exact matching are barely discussed. Considering the growing importance of techniques like Coarsened Exact Matching, this seems like a significant oversight. In addition, the book contains no exercises making it difficult to use as a textbook without some supplementary material. All in all, though, this work is a must have for those engaged in Causal Inference either academically or in the business world. Even those not making active use of these techniques might find applications to their empirical work once they understand how to properly use Propensity Score analysis.

item as expected

論理的な文章の書き方について学ぶために買いました。人工知能と会話しながら読んでいます。ルービンの文章を解剖して下さい、と依頼して、この文章でどのような論理展開が行われているか？、ここで使われている動詞や接続詞はどのような役割を果たしているか？、なぜ他の単語ではなく、この単語が選ばれているか？と聞くと詳細に説明してくれます。いちいちどの単語を使うのがベストなのかを考えるべき、ということがよくわかります。道は遠いですが、コツコツ学ぼうと思っています。