Quote:
Originally posted by Tyrone Slothrop
Spanky, re-read your own post if you need to, but it does not explain anywhere how the term "median" is ambiguous. You explain why medians and means diverge, and you explain why medians will change if you data sets change, but there's nothing ambiguous about the term.
Let's make this concrete. Here's a data series:
1, 2, 3, 9, 10.
What's the median? 3, right? What's ambiguous about that? The mean is 5, so the mean is greater than the median, but that's not a defect in the concept of medians, it's just a reflection that the two terms mean different things.
You seem to be especially combative about this because you think I was using median incomes to imply mean incomes were falling. But that's not what I said. I just said medians were falling.
|
I am not being combative. Anytime I hear the term Median my radar goes up. In his september 7th column (populist myths about income equality) which is an article that some what backs up what I am saying, the minute Brooks started referring to Medians I discounted those stats as meaningless. My guess is he started referring to median incomes because the relevent facts did not back up his argument.
http://www.nytimes.com/top/opinion/e...oks/index.html
I don't care what the situation. If I say it once, I will say it a hundred times, your definition above is how you find the Median in basic mathematics. However, in statistics, especially in statistical analysis where you use sample groups, the median is not found using that system. The term "Median" is a much more expansive term than you seem to realize.
Statisticians pick seemingly random spots on the bell curve and call it the Median. I just don't trust it when people use the term.
