Article on statistics in the media

[kylenano]

A glass half empty? (http://news.bbc.co.uk/1/hi/magazine/5021406.stm)

I heard this on Radio 4 a few days ago and was interested in what Lisa Jardine had to day. She sheds some light on how the media sensationalise statistics from dandruff shampoos to Herceptin. And when does a handful of grain become a heap?

[Gillianren]

When it's in a pile?

[mickal555]

Oh dear...

[mugaliens]

Stats themselves are capable of lying. It's only through misunderstanding, misuse, or abuse (not following the rules of statistics) that they cease representing reality and start representing the abuser's personal agenda.

[farmerjumperdon]

Stats themselves are capable of lying. It's only through misunderstanding, misuse, or abuse (not following the rules of statistics) that they cease representing reality and start representing the abuser's personal agenda.

I'm guessing you meant to say "not capable" in that 1st sentence.

My Six Sigma trainers credo:

"In God we trust. All others bring data."

[Inferno]

Another instance of this is where the news reports will claim "Crime up by 10%!" Which is, of course, a completely meaningless figure unless a trend is produced. eg crime may have gone down by 15% the previous year but that's not such an interesting headline.

I also watchout when someone says that "worst in the last 10 years", or "gone up for 7 years". Makes me think, why did they pick that start date? What happened in the year(s) before the ones they are telling you about?

[Glom]

Another instance of this is where the news reports will claim "Crime up by 10%!" Which is, of course, a completely meaningless figure unless a trend is produced. eg crime may have gone down by 15% the previous year but that's not such an interesting headline.

I also watchout when someone says that "worst in the last 10 years", or "gone up for 7 years". Makes me think, why did they pick that start date? What happened in the year(s) before the ones they are telling you about?

That's like the ozone thing. They said ozone went up in 2005, which is terrible but neglected to mention that it went up from the all-time low in 2004.

[Tbutyl]

One of my favorite uses of statistics was by a former U.S. Surgeon General. I do not remember if it was during a speech or a press conference but the topic was dealing with drug use in the U.S. and the Surgeon General made the comment “...60% of most ...”.

And before anyone else does, I might as well throw out that famous quote attributed to Mark Twain:

There are three kinds of lies: lies, damned lies and statistics.

[ToSeek]

Another instance of this is where the news reports will claim "Crime up by 10%!" Which is, of course, a completely meaningless figure unless a trend is produced. eg crime may have gone down by 15% the previous year but that's not such an interesting headline.

Or if it means that there were 11 crimes instead of last year's 10.

[Frog march]

posts by me, in this thread, just went up by ∞%.:D

[Inferno]

A related media trick is when the preceed a news story by saying "A new study indicates.....". The results of study invariably show the opposite of the generally accepted view is in fact true. Now while there may be nothing wrong with that study, it can leave alot of people to think that the new study cancels out the old.

People also often use recent studies in an attempt to back up their arguments. For example, saying "A recent study shows that cell phones can lead to cancel". They conveniently then leave out that all the past studies show no such link.

Such arguments appear to surface alot in relation to illegal drugs. You'll see a big newspaper headline: "Study shows ecstasy leads to drain damage". But studies that shows no such finding won't get a mention.

[NEOWatcher]

People also often use recent studies in an attempt to back up their arguments. For example, saying "A recent study shows that cell phones can lead to cancel". They conveniently then leave out that all the past studies show no such link.

Chocolate is good for you, it gives you energy.:rolleyes:
Such arguments appear to surface alot in relation to illegal drugs. You'll see a big newspaper headline: "Study shows ecstasy leads to drain damage". But studies that shows no such finding won't get a mention.
Cast iron or PVC drains? :think:

[Glom]

Such arguments appear to surface alot in relation to illegal drugs. You'll see a big newspaper headline: "Study shows ecstasy leads to drain damage". But studies that shows no such finding won't get a mention.

I believe term is ratchett reporting. It always goes one way, never the other.

[Matherly]

Did you know that 47% of all statistics are made up?

[NEOWatcher]

Did you know that 47% of all statistics are made up?
Did you know that joke is said in 85% of all statistics discussions?

[Matherly]

Did you know that joke is said in 85% of all statistics discussions?

Did you know that only 23% of the readers still find it funny?

(Sorry, Neo. You're right, it has been beat to death with a rock. I promise to be a good boy from now on.)

[zebo-the-fat]

I think it's shocking that 50% of school leavers are below average in mathematics!

[SeanF]

I think it's shocking that 50% of school leavers are below average in mathematics!
That would be "median," not necessarily "average." :)

[zebo-the-fat]

That would be "median," not necessarily "average." :)

Picky! :razz:

(maybe I am one of the 50%)

[HenrikOlsen]

Change to about half are below average and you got something that's easier to say and is still true.

I just read the Feynmann lecture where he mentioned the emotional problems of starting university, where about half the new students will suddenly find themselves transformed from being in the top 5% to being below average.
It's not all who can cope with that.

[mickal555]

What's the one..?

40% of sickies are tacken on mondays and fridays...

[mugaliens]

Here's a twist: "I consider myself in the top 5% of everyone I meet." "Then again, I've always considered myself above average...." (from a movie)