< Back
EMWA Logo EMWA Logo

Statistical Analysis of Binary Data

Level:
Advanced
Area:
Professional Techniques
Code:
PTA8
Duration:
03:30:00
Workshop profile:
Medical writers who wish to understand the analyses of binary data (i.e. yes/no outcomes such as death, response to treatment etc.). Participants should be familiar with basic statistical concepts such as P-values and confidence intervals.
Objective:
To gain familiarity with some of the techniques used for analysing binary data and the various measures of association, such as odds ratios or relative risks, used in analysing binary data.
Content:
The workshop will focus on understanding what the various statistical tests are, why they are done, and how to interpret the results. This will not be a tutorial in how to do the statistical tests (other than some of the simplest ones), as it is assumed that participants will generally receive the results of analyses from statisticians and be given the job of interpreting those results. • Definitions of binary data • Simple statistical tests to compare binary data in 2 treatment groups: chi-squared test and Fisher’s exact test • Measures of association of binary variables: odds ratios, risk differences, relative risks • The Mantel-Haenszel test and other more sophisticated analyses • Logistic regression: the most powerful technique for analysing binary data