Difference between Sensitivity and Statistical Power?


Could anyone point me towards an explanation of the difference between Sensitivity (as described here: http://en.wikipedia.org/wiki/Sensitivity_and_specificity) and Statistical Power?

I'd assumed the two terms were synonyms, until I read this on the above mention Wikipedia page:

"In the traditional language of statistical hypothesis testing, the sensitivity of a test is called the statistical power of the test, although the word power in that context has a more general usage that is not applicable in the present context."

Many thanks for any advice you can give!

I have not seen sensitivity used that way before. I can understand the analogy, though. I think power is used in regards to rejecting a null hypothesis. Sensitivity is just a relative frequency statistic.
Well right on the page it says "power=sensitivity" ending that debate. I would amend that to be "empirical power = sensitivity" for the related example personally.

In the page the "Positive" column corresponds to what we would think of is a true alternative hypothesis. You can see the sensitivity calculation is the number of times the null was rejected (the test result was positive) when the alternative was true (when the positive was true) divided by the sum of the times the alternative was true. ... which is power.

But since its not a analytic probability and rather an empirical observation we shouldn't call it plain power. We should call it empirical power. I suppose in the context of that page every sensitivity is an empirical sensitivity.
Last edited: