This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: 1 1 Calculation of Sample Size while Controlling False Discovery Rate with Application to Microarray Peng Liu 3/25/2008 2 Question ¡ Microarray experiments are expensive. • Cost: ~ $$$ • Effort: Experiment and data analysis take considerable time ¡ How many chips are needed for reliable results without wasting money and time? Î Estimate sample size before experiments! What is a robust and simple statistical method? 3 Calculating sample size when testing one hypothesis ¡ Type I Error : false positives ¡ Type II Error: false negatives (1-power) ¡ Power : true positives ¡ Want: sample size big enough so that we control type I error at a certain rate and achieve a desirable power if we know the true parameters. correct (Power) Type II Error False Null Type I Error correct True Null Reject Null Accept Null Hypothesis 4 Calculating sample size when testing one hypothesis ¡ Two types of errors (page 216 of handout) Type I Error : Pr( reject H |H is true) Type II Error: Pr( do not reject H |H a is true) = (1-power) ¡ Power : Pr( reject H |H a is true) (page 222 for computing power) correct (Power) Type II Error False Null Type I Error correct True Null Reject Null Accept Null Hypothesis 2 5 Hypothesis test ¡ Test is set up so that type I error rate is fixed and small. The tolerable type I error rate is often called the significance level and denoted with...
View Full Document
- Spring '08
- Statistical hypothesis testing, Type I and type II errors