CMHL1001 Chapter Notes - Chapter 7: Bonferroni Correction, Statistical Inference, Descriptive Statistics
3 Aug 2018
School
Department
Course
Professor
Document Summary
Have they determined whether their groups are comparable: most comparative clinical trials include a table or paragraph in the text showing baseline characteristics of groups being studied. What type of data have they got: numbers are often used to label properties of things, he shape of the distribution reveals results. Transforming data is not cheating it ensures that data values are given appropriate emphasis in assessing overall. It can be non-normal (skewed) that can be transformed to give a normal shape graph or can"t be transformed. effect. In order to demonstrate that a has caused b u need more than the correlation coefficient. Using tests based on the normal distribution to analyse non-normally distributed data is not good practice. Have assumptions been made about the nature and direction of causality: the presence of association between variables tells nothing about the presence of direction.
