I am currently using the 54845A oscilloscope to check Gaussianity of a noise source using the histogram setting on the scope, and I am having difficulty understanding the statistics I am getting, specifically the median value. Because this noise source follows a Gaussian distribution I would think the median would be near zero. However the median reads 7.143mV no matter how many different noise sources I test. I can change the offset of the data to force the median to zero, but this doesn't explain the problem. This value stays the same unless I change the vertical or horizontal resolutions, in which the value gets larger or smaller but still is static. So A) shouldn't this number be closer to zero? and B) shouldn't this number change when different components are tested?