Home 
Contact 
Profile 
Videos 
Prepared by Professor Omar Hasan Kasule Sr. MB ChB (MUK), MPH (Harvard), DrPH (Harvard) Chairman of the KFMC IRB
1.0 PARAMETRIC ANALYSIS OF CONTINUOUS VARIABLES:
- The Pearson correlation coefficient studies linear relation between 2 variables
- Parametric tests based on the mean are used for normally distributed data
- The t-statistic is used to compare 2 samples. It may be paired or independent.
- The F statistic (also called ANOVA) is used to compare 3 or more samples.
2.0 NON PARAMETRIC ANALYSIS OF CONTINUOUS DATA:
- Non-parametric methods based on the median are used for data not normally distributed. Virtually every parametric test has an equivalent non-parametric one.
- The Spearman is the non-parametric equivalent of the Spearman coefficient
- The signed-rank (Mann-Whitney) test is used for 2 paired sample medians.
- The rank-sum test is used for 2 independent sample medians. T
- The Kruskall-Wallis test is a 1-way test for 3 or more independent sample medians.
- The Friedman test is a 2-way test for 3 or more independent sample medians.
3.0 ANALYSIS OF DISCRETE DATA: 2 OR MORE PROPORTIONS - 1:
- An approximate method (chi-square) is used for large samples and the exact method (Fisher's Exact Method) is used for small samples.
- The data is laid out in contingency tables and is inspected manually before the application of statistical tests. The Pearson chi-square is computed based on the observed and expected frequencies of each cell in the contingency table and is in essence a measure of the deviation from the ‘average’.
4.0 ANALYSIS OF DISCRETE DATA: 2 OR MORE PROPORTIONS - 2:
- Exact methods are used instead of the chi-square test for small samples less than 20.
- Stratified and matched analysis of proportions: The Mantel-Haenszel chi-square is used to test 2 proportions in stratified data. The MacNemar chi-square is used for pair matched data.