Statistics can be a bit daunting, but the folks over at lifehacker have a nice “quick 4” to look out for. Check it out here.
One of the more interesting for us in assessment is Simpson’s Paradox. We often find that since we have different types of populations within our cohort that we have to examine the data we generate based on those subpopulations. It is a nice description of what is a common problem for us.
Finally, the closing advice that we always look for the “story” within the data is a great reminder.