Two-Sample Equivalence Test
Power and Sample Size
Power Analysis - Power
|
|
Two-Sample Equivalence TestPower and Sample Size |
|
For a 2-sample equivalence test, power is the ability to establish that the test mean is equivalent to the reference mean. In other words, power is the probability that you will conclude that the difference between the means is within your equivalence limits, when this is in fact true. See Power for equivalence tests for further discussion.
If you specify the sample size and the difference that you want to accommodate and still be able to claim equivalence, Minitab calculates the power for the test.
Example Output |
|
|
2-Sample Equivalence Test
Power for difference: Test mean - reference mean Null hypothesis: Difference ≤ -1 or Difference ≥ 1 Alternative hypothesis: -1 < Difference < 1 α level: 0.05 Assumed standard deviation: 0.41
Sample Difference Size Power 0.0 8 0.997132 0.2 8 0.980240 0.4 8 0.872057 0.6 8 0.583518 0.8 8 0.236964
The sample size is for each group. |
Interpretation |
|
The pain reliever analysis shows that, if the difference is larger, then you have less power to claim equivalence. If the difference is 0, then the power for the test with 8 observations in each group is approximately 0.997. If the difference is as much as 0.8, then the power is approximately 0.24.
With a power of 0.24, the chance that you will conclude that the difference is within your equivalence limits, when this is in fact true, is only 24%. Typically, you want a power of 0.9 or more.
The power curve is a useful way to visualize the relationship between power and the difference.
Free online chart generators for statistical analysis:
Useful tools:
