posted by Mike Madison at 7:49 PM /
mtl magazine brings information. See if you can figure out what it means.
In the March 2009 issue, "Mt. Lebanon By the Numbers" reports that the "Mean family income" in Mt. Lebanon is $127,605. What's the source? "Numbers are based on regional data and trend analysis." See page 25.
Not long ago, MTL magazine distributed a profile of Mt. Lebanon to prospective advertisers that included this line about Mt. Lebanon: "The median family income is $98,042." No source was provided. I wrote about the profile in an earlier post.
Statistics wizards in the audience will recognize immediately that mean and median are different things. Where the median is lower than the mean for the same collection of data, then you are likely to find a small number of high-end outliers. In other words, the high mean family income in Mt. Lebanon doesn't mean that we all make an average of $127,000. It means that most families make less than that, and a relatively small number of families make much more.
But there is more.
According to the 2000 census, Mt. Lebanon's median household income was $60,783. Its median family income was $79,744. I posted about these statistics more than two years ago. The census data is available here.
What's going on? I assume that all of these numbers are accurate, or as accurate as they can be under the circumstances. (It is always possible that mtl's sources, whoever they are, have engaged in a bit of wishful thinking.)
[I deleted the rest of the original text of this post; Jefferson Provost's comment, which critiqued it, suggests that I should stick to reporting what I read rather than trying my hand at quantitative analysis!]
Maybe the Mt. Lebanon "Bubble" means something other than what we all thought it meant.
In the March 2009 issue, "Mt. Lebanon By the Numbers" reports that the "Mean family income" in Mt. Lebanon is $127,605. What's the source? "Numbers are based on regional data and trend analysis." See page 25.
Not long ago, MTL magazine distributed a profile of Mt. Lebanon to prospective advertisers that included this line about Mt. Lebanon: "The median family income is $98,042." No source was provided. I wrote about the profile in an earlier post.
Statistics wizards in the audience will recognize immediately that mean and median are different things. Where the median is lower than the mean for the same collection of data, then you are likely to find a small number of high-end outliers. In other words, the high mean family income in Mt. Lebanon doesn't mean that we all make an average of $127,000. It means that most families make less than that, and a relatively small number of families make much more.
But there is more.
According to the 2000 census, Mt. Lebanon's median household income was $60,783. Its median family income was $79,744. I posted about these statistics more than two years ago. The census data is available here.
What's going on? I assume that all of these numbers are accurate, or as accurate as they can be under the circumstances. (It is always possible that mtl's sources, whoever they are, have engaged in a bit of wishful thinking.)
[I deleted the rest of the original text of this post; Jefferson Provost's comment, which critiqued it, suggests that I should stick to reporting what I read rather than trying my hand at quantitative analysis!]
Maybe the Mt. Lebanon "Bubble" means something other than what we all thought it meant.
2 Comments:
I think you have it backwards. If Bill Gates moved to Lebo, the mean income would go up a lot, but the median would barely change. (The median of 1, 2, and 3 is 2, the median of 1, 2, 3, and 1000 is 2.5)
Also WTF does "trend analysis" have to do with computing the mean income? Computing the mean family income is pretty straightforward: add up all the family incomes, and divide by the number of families. Getting those data might not be so straightforward, but if they're using some kind of trend analysis then they're not really reporting the mean, but some kind of estimate. It's impossible to know how accurate that estimate is without knowing what kind of trend model they're using and how it was computed. Extrapolating from trends is a risky business, and extrapolating growth trends is especially risky, as anyone who reads the national news knows by now.
The next census is next year. Anyone want to take bets on how far off the census figures are from the MTL figures, and in which direction?
You're a math guy, so I'll trust your numbers. I'll fix the post.
Post a Comment
<< Home