Just a quick vocabulary note by way of blog entry today.
In statistics, the word "power" is used as a synonym with "sensitivity". The test of a hypothesis has POWER to the extent that it is likely to detect an effect if it exists.
Typically in scientific experiments there will be a null hypothesis and an alternative hypothesis of this form: the null is that there is no relationship between two variables, the alternative is that there is one.
If a powerful test fails to give the tested-for result then, by definition, it supports the null hypothesis.
As you may already have determined, I continue to make my way slowly through the book by Deborah Mayo that I have mentioned here a couple of times. Boning up on the vocabulary of the field as I go.
The above diagram should help explain the significance of power so understood. The further away the modes of the two bell curves; the smaller the possibility of an erroneous decision.
Comments
Post a Comment