The power of a test is its ability to detect an effect. It is always possible that, due to sampling error, a test will lead you to the wrong conclusion. Assessing power allows you to determine the probability that the test will correctly identify an effect if one exists.

If a test has low power, you may fail to detect an effect and mistakenly conclude that none exists. If the power of your test is too high, very small and possibly uninteresting effects can become significant.

If you provide the difference that you want to be able to detect and the size of your sample, Minitab will calculate the power of the test.

For the energy data, the researcher suspects that the mean energy expenditure for upper-income households may be different from the published value of $1080. He is interested in detecting a difference of $100 or more. (He considers an effect smaller than this to be of little interest.) He wants to know how powerful his test will be if he samples 15 households.

The results indicate that the test has a power value of 0.504247. Thus, if the mean energy expenditure for the population is $980 or $1180 instead of $1080, there is only a 50.4247% chance that the test will detect this.

Such ambiguous results may not be worth the effort of conducting the test. Obviously, it would be desirable to increase the power of the test so that he can have more confidence in the results.