Top critical review
6 people found this helpful
Thorough but not very satisfying
on September 15, 2014
This is a thoroughly-written book. On one hand I was very glad that it didn't turn out to be a pop-economics type of book. On the other hand, I think the book is largely meant for a more academic audience, so I found it very dense to get through. It describes the problem very clearly, and gives a detailed account of the history behind the statistics, which was interesting.
What I wish the book had, however, was more help for people who want a way out! It spends 90% of time talking about the problem of statistical significance and the history behind it, but I was already in agreement with them so I didn't need any convincing.
I was hoping for more guidance on alternative approaches, or at least more detail on Gosset's thinking and ideas. They make vague references to loss functions, power analysis, etc. as much better approaches, but if you don't know very much about those things you're pretty much on your own to read something else.