The book of Shipley provides an easy to read introduction in the field of structural equations and causal inference from experimental data. Most concepts are very well explained and examples are provided to see how to apply the concepts practically. Without any reservation I recommend it to none experts in this field who want to learn the intermediate basics directly applicable to own problems, e.g., for biologists.

However, for people trained in more theoretical fields, e.g., computer science, the thorough verbal explanations of mathematical formular (which are provided) is a little tiring after the first 100 pages, because the relative simple formulas can easily be interpreted. Instead, I wished the book would provide more alternative approaches in form of statistical test to test for causality. For this reason, I would recommend not to rewrite the complete book, but only include one additional chapter written in a more technical style. Then it would be close to perfect.

Finally, I want to remark, that the historical notes given throughout the book are not only very intesting but also inspiring, because they remind not to oversee mathematical developments in the none mathematical literature. SEM is certainly a good example.

To summarize, this is really a decent book!

Скачать книгу бесплатно (djvu, 2.92 Mb)

---

Читать «Cause and Correlation in Biology: A User's Guide to Path Analysis, Structural Equations and Causal Inference»