Include current observation in standard deviation calculation?

Question

This isn't a question about execution, more theory. I'm monitoring data flow daily and want to calculate how many standard deviations from the mean today's value is compared to history. My question is, should I include today's value when calculating the standard deviation and mean? Or exclude it?

Answer 1

Well, the question you are trying to answer is, is the current reading "unusual" considering the a collection of "usual" measurements. I think that only makes sense if the current reading is excluded from what's considered "usual"; otherwise you have some slightly circular reasoning going on.