Tools and techniques - Statistics: Analysis of continuous data using the t-test and ANOVA

摘要

In clinical research, we often examine continuous variables such as blood pressure, ejection fraction, laboratory values (e.g., cholesterol), and angiographic variables (e.g., percent stenosis). We may, for example, want to compare these measurements between different patients or between different time points. To compare continuous data, parametric or non-parametric tests of significance may be applied. Which of these tests is appropriate depends on several factors, including the nature of the data to be analysed. These data should meet the assumptions that are required for the particular test. This paper describes parametric methods and the circumstances under which these methods are appropriate. Its aim is to give a general overview, without claiming completeness; hence, exceptions are possible in specific situations.