Statistics Book

Decent textbook on statistics. The book itself is aimed at both a general public and undergraduates, meaning that the content is not too sophisticated. I read this to freshen up my rusty statistical knowledge (to prepare for some tests), so in this sense the book has served its purpose.

The authors write in an accessibly style and use lots of simple examples to lay down the foundations of the most important statistical theories and tools. Descriptive statistics, inferences, frequency theory of chance, using models, and tests of signicifance like the chi square test (for testing frequencies) and t-tests (both between-group and within-group analysis).

It would have been nice to read some more about the historical context of these ideas and their development over time. Also, it would have been nice to read some more about the philosophical problems surrounding statistics as tool for reliable knowledge. The authors mention that in modern science, especially the alfa sciences, statistics is over-used and in many cases merely used as weighty language to proffer up results.

I can attest to this last point - any analysis based on models is simply based on the assumptions of the model itself. Trash in = trash out. Modern sciences like psychology like to present their results as significant, but they're really only presenting their own assumptions in weighty language. (It is common knowledge that replicability, for example, is a huge issue in psychology and that sociological research is heavily influenced by political-ideological assumptions). Would have been nice to read some famous case studies...

Anyway, the book is decent - I can't really recommend it besides digging up old knowledge. To learn statistics one needs some more guidance, preferably within a course setting.