Mar 12th 2017, 11:15 AM
Join Date: Jun 2016
unless your 6 samples have a relationship that can be defined relative to each other on a quantifiable axis (rather than just sample number), I cannot see that you can define a single standard deviation for the final graph.
However, you can derive standard deviations individually for each sample and then derive the average and the standard deviation of these 6 standard deviations.
This will give you a measure of the accuracy and repeatability of your experiment (although 6 samples is a a bit small for deriving a good standard deviation).