Page 1 of 1

News usage of statistics

PostPosted: Tue Oct 17, 2006 11:18 am
by Ivan
Anyone ever cared to think about some of the statistics shown in the news?

Following the credo 'Never trust statistics you haven;t manipluated youself' you can get assurance for almost ANYTHING which links demographics to things like crime rate, education e.a.

Lately there were some nice examples of this one...

NATO about afghanistan
A recent NATO inquiry showed that a probable 70% of the population did support the Taliban. which resulted in some theatrical 'major concern' press conferences.
Actually it is one of those 'No Sh!t Sherlock' finds... using the following facts.
1: With the literacy rate up there i see the whole inquiry as something fictional to begin with... If they really did an printed inquiry the results would be way different.
2: If you look up the population buildup, its quite clear what the meaning of 'Taliban' in the context of the inquiry was: Anyone not supporting the current government (that includes the Hekmatyr supporters who in reality DON'T EVEN SUPPORT the Taliban)
3: Dont forget that a lot of these guys haven't got the slightest grasp at the concept 'nation'. Loyalty ends with either tribe or language group.

Anyway... anybody with a map and some common sense could have come up with these numbers, without even leaving his desk.

And the second one which involves a lot of 'be really scared' talk.

In holland there is a guy who has to write something sensible about 'turrism' every three months, mainly to keep the majority of the population scared enough to swallow any privacy restriction measure the government comes up with.

As he needs some variation to keep his credibility up he decided to choose a different minority group than the usual one (turks instead of north africans).
Anyway the outcome was that the dutch should be scared... (but that is where he gets paid for)

And here come the arguments to send this report single flight to the dustbin.  While it is likely that there was some resarch behind this one, given the buildup of the resarched population the results are both obvious and largely overrated.

1: A large part of the Turks in Holland are actually Kurds, some of which obviously supporting the KDP (ex PKK, both are on the EU terrorism list). It looks a lot more scary than it actually is given the fact that the Iraqi section of the same movement is big friends with the USA...
2: Ever heard of Mili Gorus... it's the major backing group behind the current Turkish prime minister... It has been banned in Germany for their links with some extremist nationalistic groups and is on the EU watch list. Apart from their intimidating looking mass rallies the majority isnt that dangerous.
3: With those two largely skipped there are a few small groups left that really pose a danger... but by any means not the number the report shows

So next time you see some TV guy wave another 'x precent supports cause y' report, take a look at the numbers before you take it as the truth... most of their finds are just hot air, treated in such a way that the average viewer takes them for granted...

Re: News usage of statistics

PostPosted: Tue Oct 17, 2006 2:58 pm
by pepper_airborne
I learned to never trust statistics in school :).

Re: News usage of statistics

PostPosted: Tue Oct 17, 2006 3:46 pm
by Chris_F
There's an election coming up in the US which means the politicians are out in full force.  One state to the south of me there's a ballot refferendum to allow the sale of wine in grocery stores.  There is an anti-campaign in place (probably funded by the liquor stores) which is using some fool statistic like: states which allow the sale of wine in liquor stores have 9x the drunk driving fatalities as Massachusetts.

My guess is the truth is that this is the worst case.  And if you look at the worst case of states that do allow wine sales in groceries the fatality rate is probably similar.  And I'd even bet that on average there is no discernable difference between states which do and don't sell wine in groceries.

But in a 30 second ad that type of stats quote can be very persuasive.

Did I mention I hate politics?

Re: News usage of statistics

PostPosted: Tue Oct 17, 2006 11:53 pm
by beefhole
Psychology 101.  Don't trust stats you haven't collected yourself or that weren't collected in a very scientific, non-misleading manner.

Re: News usage of statistics

PostPosted: Wed Oct 18, 2006 12:33 am
by llamedos
82.75% of statistics are made up on the spot  ;D ;D

sir Martin

Re: News usage of statistics

PostPosted: Wed Oct 18, 2006 12:41 am
by BAW0343
82.75% of statistics are made up on the spot  ;D ;D

sir Martin



79.34% of that post is Bull  ;D

Re: News usage of statistics

PostPosted: Wed Oct 18, 2006 1:35 am
by Ivan
[quote]There's an election coming up in the US which means the politicians are out in full force.

Re: News usage of statistics

PostPosted: Wed Oct 18, 2006 6:16 am
by beaky
"Facts are stubborn, but statistics are more pliable."

-Mark Twain

Re: News usage of statistics

PostPosted: Wed Oct 18, 2006 6:53 am
by Chris_F
3 out of 4 people make up 75% of the population.

Re: News usage of statistics

PostPosted: Wed Oct 18, 2006 10:54 am
by beaky
More comment on this from Mark Twain (a regular cornucopia of wry observations):

'By further figuring, it appeared that between New York and Rochester the Erie ran eight passenger-trains each way every day - 16 altogether; and carried a daily average of 6,000 persons. That is about a million in six months - the population of New York City. Well, the Erie kills from 13 to 23 persons out of its million in six months; and in the same time 13,000 of New York's million die in their beds! My flesh crept, my hair stood on end. "This is appalling!" I said. "The danger isn't in traveling by rail, but in trusting to those deadly beds. I will never sleep in a bed again." '


It's an excerpt from a little essay called "The Danger of Lying in Bed"; interesting in light of today's hand-wringing over air safety (aside from terrorism concerns), as it's about railroad fatality headlines  and accident insurance from the 1870s.
Read the whole thing here:

http://www.druglibrary.org/schaffer/general/twins2.htm

Re: News usage of statistics

PostPosted: Wed Oct 18, 2006 2:13 pm
by born_2_fly


79.34% of that post is Bull  ;D



Approx 89.576% of that post was a lie!  ;D ;D ;D ;D ;D

Re: News usage of statistics

PostPosted: Wed Oct 18, 2006 2:24 pm
by Fozzer
Only 2% of me is studying this post...

The other 98% is wondering what's for Supper...

Paul...just wondering if Stat Is a Stick...?

LOL... ;D...!