Sample size calculation: Area under ROC curve
Calculates the required sample size for the comparison of the area under a ROC curve with a null hypothesis value. The sample size takes into account the required significance level and power of the test (see Sampling: Introduction).
You want to show that the AUC of 0.725 for a particular test is significant from the null hypothesis value 0.5 (meaning no discriminating power), then you enter 0.725 for Area under ROC curve and 0.5 for Null Hypothesis value. You expect to include twice as many negative cases than positive cases, so for the Ratio of sample sizes in negative / positive groups you enter 2.
For α-level you select 0.05 and for β-level you select 0.20 (power is 80%).
After you click the Calculate button the program displays the required sample size.
In the example 19 cases are required in the positive group and 38 in the negative group, giving a total of 57 cases.
A table shows the required sample size for different Type I and Type II Error levels.