A metallurgist wants to repeat a particular experiment with four different treat

Question

A metallurgist wants to repeat a particular experiment with four different treatments and wants the design to be sensitive enough so that it is likely to detect a difference of 10 between treatment means at the 5% level. He assumes that the error standard deviation will be about the same as the value of 6.726 calculated in the previous experiment. The following output from a statistical software package shows the result of a power calculation for an experiment with five observations per treatment. What is the power? What recommendation would you give the metallurgist regarding the usefulness of this proposed experiment? One-way ANOVA Alpha = 0.05 Assumed standard deviation = 6.726 Number of Levels = 4 The sample size is for each level. A. The power is 0.387777. Do not run the experiment. B. The power is 0.387777. Run the experiment.

Explanation / Answer

The power is 0.387777 . We do not run the experiment. Power should be greater than 0.7

Related Questions

Navigate

• Browse (All)
• Browse A
• Subjects

---