P-Value Adjustment Strategies for Multiple Statistical Testing

About this Class

When several statistical tests are performed, simultaneous use of the resultant “raw” (i.e., test-generated) P-values can lead to an increased overall false-positive rate. To keep that rate under control, various statistical procedures have been developed. These procedures, generally known as P-value adjustment strategies, are the main subject of this lecture. We will review the conceptual foundation behind P-value adjustment, take a closer look at some common adjustment strategies, and see how P-value adjustment can be performed in practical situations relevant to biomedical research. Beginner level of statistical knowledge is expected, intermediate is preferred.