Fred:

Being the typical engineer (we have to take a lot of statistics courses) I like a lot of data points to make sure I'm not looking at an "outlier" or erroneous data point. I typically try to use 10 repeats of the same experiment (if it is not too costly... Login to read more.