Am höchsten bewertete kritische Rezension
Eine Person fand diese Informationen hilfreich
Does what it should, minor errors
am 20. Januar 2017
The book gives a very good overview of how to achieve certain tasks in R. It's good for beginners and good to look things up.
However, it does not cover everything, for instance time-formatting. Neither does it cover the mathmatical theory behind the functions.
The worst part is the statement on the back cover by JD Long: "With 95% confidence I fail to reject..."
1. If a statistical test rejects a Hypothesis, one may have 1-alpha (e.g. alpha = 5%) confidence, that the hypothesis is wrong, because the test is designed this way. The error-percentage alpha is called first order error and is controlled by calculating the according statistic and quantiles.
2. If you fail to reject the hypothesis you may (A) be right or (B) make a second order error, which cannot be calculated, hence JD Long's statement is NONSENSE!!! Some randomised tests try minimizing the error, but even then it cannot be controlled or calculated.
So please do not believe everything you read!