Ehh, good queston. You're starting earlier than I did, with what really is a foundation issue - how words can describe statistics or vice versa.jimmie said:I agree with that statement.
Maybe a little off topic, but the next question was: what makes statistics relevant, so as to be represented by invented words such as "poverty"?
The word "poverty" predates modern statistical analysis, so to me it is clear that if one chooses to use the word, they should make the statistics match the definition of the word as opposed to changing the definition of the word to match the statistics. You can't just grab any random statistics and attach the label "poverty". Otherwise, what stops me from attaching the word "poverty" to the ratio of gray hairs on a person's head? Below 20% gray and you are "rich" and above that you are "poor"? Meaningless, right? So too with misapplying the word "poverty" to what is actually just "income distribution".
The point is, the verbal definition of poverty is clear - and when pressed, pretty much everyone will agree on it. So it should be equally clear that any statistical analysis should be an honest attempt to measure the concept described in that definition.