Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I was just wondering about a very related problem. I do 5 measurenments of some random variable (let's say execution time) and average them. How should I report the variability of that average?

State the sample size and standard deviation?



If it were 10 or more measurements, I'd use average, standard deviation and sample size if we can expect that variable to be normally distributed.

I don't know how to make meaningful statistics with fewer data points.


Citing 10 measurements in a table or plot is probably easier than explaining how and why you are taking mean and standard deviation (or anything else for that matter...)


90% of the data fall between this point and this point. 50% of the data fall between this point and this point. etc.

MAD is more accurate than Standard Deviation to me, if you ask me.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: