I think this is a statistical fallacy...Does it have a name?

[pimp_named_ak]

This page (http://moneycentral.msn.com/content/invest/extra/P148783.asp?GT1=8007) purports to list the ten richest U.S. counties. In fact, it lists the counties which have the most millionaires in them. The introduction cleverly labels these counties as 'millionaire-congested.' This leads me to a two part question:
1. shouldn't you be looking at number of millionaires (or some other measure of wealth) per capita!?
2. assuming that the msn people (who published this page) are in fact, committing an error, does this error have a name? It seems like the opposite of ecological fallacy. (http://en.wikipedia.org/wiki/Ecological_fallacy)

[billygrippo]

[ QUOTE ]
This page (http://moneycentral.msn.com/content/invest/extra/P148783.asp?GT1=8007) purports to list the ten richest U.S. counties. In fact, it lists the counties which have the most millionaires in them. The introduction cleverly labels these counties as 'millionaire-congested.' This leads me to a two part question:
1. shouldn't you be looking at number of millionaires (or some other measure of wealth) per capita!?
2. assuming that the msn people (who published this page) are in fact, committing an error, does this error have a name? It seems like the opposite of ecological fallacy. (http://en.wikipedia.org/wiki/Ecological_fallacy)

[/ QUOTE ]

perhapse the counties are quite low population, full of the super rich. like gated comunities.

[SkinnyPuppy]

well at least now i know where to go to steal cars when i go broke! /images/graemlins/crazy.gif mmmmwwwhhhaahhhaa

[Silent A]

[ QUOTE ]
[ QUOTE ]
This page (http://moneycentral.msn.com/content/invest/extra/P148783.asp?GT1=8007) purports to list the ten richest U.S. counties. In fact, it lists the counties which have the most millionaires in them. The introduction cleverly labels these counties as 'millionaire-congested.' This leads me to a two part question:
1. shouldn't you be looking at number of millionaires (or some other measure of wealth) per capita!?
2. assuming that the msn people (who published this page) are in fact, committing an error, does this error have a name? It seems like the opposite of ecological fallacy. (http://en.wikipedia.org/wiki/Ecological_fallacy)

[/ QUOTE ]

perhapse the counties are quite low population, full of the super rich. like gated comunities.

[/ QUOTE ]

They're not. I'm not that familiar with US counties, but many of them are basically the rich parts of the US's biggest cities, or the entire city itself.

If you compare the "% of state's millionares" with US census' % of state's population you get:

county..............% of mill....% of pop
--------------------------------------------
Los Angeles, CA .... 23% ....... 28%
Cook, IL ............... 41% ....... 42%
Orange, CA .......... 10% ....... 8%
Maricopa, AZ ........ 62% ....... 61%
San Diego, CA ....... 9% ....... 8%
Harris, TX ............. 17% ....... 16%
Nassau, NY ............ 13% ....... 7%
Santa Clara, CA ...... 7% ....... 5%
Palm Beach, FL ....... 12% ....... 7%
Middlesex, MA ........ 28% .......23%

The only counties significantly richer then their state's average are Palm Beach and Nassau. In fact, the bottom 4 are probably significantly richer than the top 6.

I think this is an example of the "journalist doesn't understand anything about statistics" fallacy.

[FlFishOn]

"I think this is an example of the "journalist doesn't understand anything about statistics" fallacy. "

This is likely the same reporter/editor combination that writes million when it's actually billion and has not enough numeracy to understand or spot the difference. J school has no math class, clearly.

[guesswest]

It's still an ecological fallacy, just a different way of arriving at one.

But the flaws involved in presenting those stats with that claim attached to them are so numerous that you could grow old listing them - it's basically junk science.

[nebben]

I think its the baseline fallacy, ignoring the base rate of the statistic

[guesswest]

Wouldn't the baseline be $0 though? as oppose to the mean? I'm asking not asserting, I'm just having trouble reconciling that idea with this scenario - probably due to me sucking at statistics.

I'd suggested it remained the ecological fallacy because my understanding was that this entailed disregarding the relationship between the mean and individual variations in a model, where one doesn't reasonably indicate the other - assuming it was essentialy the same fallacy involved in representing the mean out of context of significant variations, or representing those variations out of context from the mean.

[purnell]

[ QUOTE ]
1. shouldn't you be looking at number of millionaires (or some other measure of wealth) per capita!?

[/ QUOTE ]

That would be a more meaningful statistic, but they are only claiming that the listed counties have the most millionaires in absolute terms. The only error I see is that of the reader who infers more meaning than is there.

[Marko Schmarko]

I'm not sure one could say that a road were more "car congested" if it had 5000 cars over 300 miles than one with 3000 over 50.

[purnell]

[ QUOTE ]
I'm not sure one could say that a road were more "car congested" if it had 5000 cars over 300 miles than one with 3000 over 50.

[/ QUOTE ]

Agreed. The authors do imply that their numbers are more meaningful than they are, and alot of readers will swallow it whole.

[guesswest]

[ QUOTE ]
[ QUOTE ]
1. shouldn't you be looking at number of millionaires (or some other measure of wealth) per capita!?

[/ QUOTE ]

That would be a more meaningful statistic, but they are only claiming that the listed counties have the most millionaires in absolute terms. The only error I see is that of the reader who infers more meaning than is there.

[/ QUOTE ]

I'm not sure I agree with that. The title of the article is 'The 10 Richest US Counties', which is an insane claim to attach to that statistic. Granted someone of reasonable intelligence, which is everyone on this thread, can see through that - but we shouldn't have to catch them out.

It's godawful journalism. The article is IMO clearly written with the intention that the reader infers much more than is there, whether by design or ineptitude I don't know (I suspect both) - and to make matters worse it's in their financial section where you'd think they might know better.

[bunny]

[ QUOTE ]
... and to make matters worse it's in their financial section where you'd think they might know better.

[/ QUOTE ]
It's a minor point, but in my experience financial knowledge doesnt correlate with increased numeracy skills at all. Statistical and probabilistic ignorance and misconceptions are very close to universal.

[cambraceres]

This is a problem with reporting in general. News sources would be more reputable and staffed with those of at least reasonable intelligence were the system within which they function conducive to such things.

Cambraceres

[purnell]

[ QUOTE ]
I'm not sure one could say that a road were more "car congested" if it had 5000 cars over 300 miles than one with 3000 over 50.

[/ QUOTE ]

One could reasonably interpret "millionaire-congested" to mean "having a large number of millionaires per square mile".

[RJT]

I’m in favor of naming it “The O’Reilly Fallacy”. The idea that one can state something as true (or opinion in his case much of the time) with there being no opportunity (he never gives his guests time to speak) to debate the statement(s).

Another idea is to name it after that poster who posted his friends e-mail here on SMP.

Like a lot of “journalism”, it is shadow boxing.