26.36 Type II error rate (\(β\))

A type II error is a false-negative difference.
Detecting no difference when there is a difference
The type II error rate can be calculated given the type I error rate and the sample size (for most experiments).
The conventional value for the type II error rate is 0.2.
- (ie it is acceptable to conclude a false-negative difference that occurs by pure chance 20% of the time in an attempt to detect true-negative differences)
The power is \(1-\beta\), or the probability of detecting a true-positive difference. The experimenter chooses the false-negative rate or power (usually 80%).