A client comes to you to ask how many responses he should take for each of 6 treatments. They are to be compared using a completely randomised design. He plans to use a significance level of a=0.05.

Suppose that he wants to be able to detect a maximum difference of 0.6 times the standard deviation with a power of at least 0.85. What is the minimum number of observations that he needs to take in total?

no idea how to start this...