Examples of pitfalls in statistical analysis--4: what is a type Ii error?
Nagata, O.
MASUI. Japanese Journal of Anesthesiology 45(12): 1473-1479
1996
ISSN/ISBN: 0021-4892 PMID: 8997049 Document Number: 459087
When assessing the results of statistical analysis, we have to consider two types of errors, i.e., type I errors, which present the risk of erroneously judging a true hypothesis to be false, and type II errors, which present the risk of erroneously judging a true hypothesis to be true. If we use only type I errors to detect the significances in statistical analysis, we can not avoid making the erroneous judgement that a true hypothesis is false. When the results of analysis are not significant, we have to calculate type II errors and discuss the probabilities of having missed the truth. In this article, I present an example to illustrate two types of errors which can occur in assessing the result of statistical analysis.