statistics review 3 hypothesis testing and p values统计评估3假设检验和p值.pdf

Critical Care June Vol 6 No 3 Whitley and Ball Review Statistics review 3: Hypothesis testing and P values Elise Whitley1 and Jonathan Ball2 1Lecturer in Medical Statistics, University of Bristol, Bristol, UK 2Lecturer in Intensive Care Medicine, St George’s Hospital Medical School, London, UK Correspondence: Editorial Office, Critical Care, editorial@ Published online: 18 March 2002 Critical Care 2002, 6:222-225 © 2002 BioMed Central Ltd (Print ISSN 1364-8535; Online ISSN 1466-609X) Abstract The present review introduces the general philosophy behind hypothesis (significance) testing and calculation of P values. Guidelines for the interpretation of P values are also provided in the context of a published example, along with some of the common pitfalls. Examples of specific statistical tests will be covered in future reviews. Keywords hypothesis testing, null hypothesis, P value The previous review in this series described how to use confi- nitrate on mortality, sampling variation means that it is dence intervals to draw inferences about a population from a extremely unlikely that exactly the same proportion of patients representative sample. A common next step in data analysis in each group will die. Thus, any observed difference is calculation of P values, also known as hypothesis testing. between the two groups may be due to the treatment or it Hypothesis testing is generally used when some comparison may simply be a coincidence, in other words due to chance. is to be made. This comparison may be a single observed The aim of hypothesis testing is to establish which of thes