Test for differences in variance?
The goal of the study is two compare two devices used for imaging (say,A and B). It involves taking one picture with each device and measuring an area. The area is computed three times for each picture, the idea being that the variance in the measures is lesser in one photograph compared to the other, thus demonstrating that one device is better than the other. What is the sample size required to detect a difference of 0.03 SDs, if the distribution of measures one device is mean=0.2 and SD=0.2?
