## Question 327:

1## Answer:

No answer provided yet.You need the population standard deviation which is usually not known, so you need to estimate it from sample data. If possible try sampling a few 5-10 from your experiment to estimate the standard deviation. See if there is historical data available. If that's not possible, if you can estimate the likely range of the data, divide that range by 6 as an estimate of the standard deviation. If you have the standard deviation and a fixed sample size as well as power, the mean difference is automatically calculated.

Mean differences are context dependent, so you need to think about how big of a difference is meaningful. Is a small difference enough or should the difference be large. Cohen has some rules of thumb which might help.