Reasoning under uncertainty
Not a set of formulas — a way of deciding how much to trust a number.
Statistics for Everyday Life · Lesson 1
Not a set of formulas — a way of deciding how much to trust a number.
Almost every statistical error is in the leap from measured to unmeasured.
“The mean score in this class was 65.”
Does this summarize what was measured, or claim something about people who were not?
The leap from measured to unmeasured is where statistics gets both powerful and risky.
If gains concentrate at the top, the mean rises while the middle does not move.
Two sides of an argument, both citing real data, disagreeing about which summary answers the question.
Think like a Maester
Before doing anything with a number, say in plain language what it claims and about whom. Most errors are visible at that step.
Mark this lesson complete to track your progress.