1. Hat Wearers Wear Hats

Statistics

The SELF SELECTING SAMPLE.

(Thanks Ivy for the No 1 rocket.)

Newspaper and magazine editors urge their readers to ‘click-on our website poll’ and then they publish the results in the next issue. The newspaper may learn about their readership. This is useful information for marketing but otherwise useless. It’s like asking hat wearers if they wear hats. Let me guess the answer? D’uh!

Included here are some results of two self-selecting surveys, which not only reveal the standard useless statistics but also some highly questionable numerical outcomes. In the Esquire Magazine Survey of Drinking (Sept 2010)  82% of their readers, who were willing to answer a survey about their drinking habits (Whereby, for some reason. 1 beer = 2 drinks), have a University Degree or higher (Or, maybe, 50% of them lie!!!!) and in the Health and Fitness (Oct 2010) magazine survey  – Guess what? – magically the numbers for all options add up to 100%. Neat! Didn’t anyone fit more than one category? (Assuming all readers of Health and Fitness mag who are bothered enough to answer a survey on fitness do some exercise.)

2. My Mum Says Survey

Statistics

The SMALL SAMPLE error.

A survey is taken but the number of people surveyed is so small as to be irrelevant; not much better than simply asking your mum for her opinion and publishing the results.

Included here is a full-page Women’s Weekly (Oct 2010) ad for an Elizabeth Arden Anti-wrinkle Cream. Look at these wonderful statics. 92%… 85% …Wow! Look at the language. Gives eyes a ‘radiant and luminous look’ Sounds like the DEVIL!!!! Read the small print.

The survey was based on 30 participants and ‘results may vary’.

                            

3. The Trouble with Double

Percentages

When the Numbers involved are so SMALL the % Stated is Meaningless.

Newspapers often state that a cancer rate has doubled or increased (See pic) by 28%. Those % changes can be meaningless. For instance, double nothing is still nothing. You need the actual numbers.

 

Here is a statistic taken from Men’s Health magazine (March 2010). According to the government funded Australian Institute of Health and Welfare the actual number of Australian males who presented with melanoma in 2005 was 6,044

or   0.549  in 1,000

or 1 in 2,000.

If these numbers increase by 28% the number of Australian men presenting with melanoma will be:

 

0.703 in 1,000

or  ~ 2 in 3,000.

These numbers are not so alarming. Then again would you take any notice of statistics of a magazine that suggests a ‘sonic boom’ from a golf club is causing deafness!!!!!

4. NO. 26 It’s Your Turn Tonight

 Probability.                                          

Independent Events are not RELATED.

Lotto draws are, as with coin flipping, Independent or UNRELATED EVENTS. Newspapers often publish the least drawn numbers prior to a major prize draw.

When on June 5 2008 the Powerball jackpot reached over $50 million, making it the biggest prize ever offered in any Australian lottery game at the time media commentators went crazy. ( See Crazy Lotto Lovers Go Bananas Again!)

The Today show on Ninesmsn website advised ‘The most-drawn Powerball numbers are 26, 22, 5, 39, 24 and 34. The least drawn numbers are 41, 32, 10, 43, 35 and 20.’ This information is worthless. These facts imply that the balls know whose turn it is and then can organise themselves so that those balls drop down the shoot. As if! The draw is random. Any number is possible.

The winning numbers, in drawn order, were: 5, 21, 11, 38 and 2, with the Powerball 33, with the final prize of $58,737,207.41

Poor old 41, 32, 10,43, 35 and 20 will have to wait for another turn!!!!!!!

5. Made UP Maths

Faux Algorithms/ Dodgy Calculations:

Guess Work disguised as a Mathematical method or equation.

Moderately reputable institutions often present statistics and/or calculations, which appear sound but really involve, at best, an educated guess; at worst, a pull-the-number-out-of-your-hat trick.

The RICH LIST is one example. Forbes is a respectable magazine but they are not privy to all the complex financial interests of various list members. How  rich is Scrooge McDuck? Like, rooolly rich, dude!!!!! So which one on the Rich List is Scrooge????

 

Then there are the boffins who produce rubbish formulae for pouring beer, making the perfect piece of toast and popping champagne. These formulae are often sponsored by manufacturers to promote a product such as a new beer brand. Some examples include The Perfect Sitcom (quality = (rd+v)f÷a+s) to promote UKTV Gold; The Perfect Joke (x = (fl + no)/p) to promote some comedian; The Perfect Day (quality = O + NS + Cpm÷T + He) to promote ice cream; The Perfect Rugby Kick (KP = CSP – s + w + r + yn + cr + sc + mt + xn + ctw), which somehow has something to do with a research company called Qinetiq; The Perfect Chip (Tesco)”  and so on. This is rubbish maths because most of the ‘variables’ ( x, t, w, etc)  cannot be measured. It’s all guesswork!!!! See Mathspig Post Britney’s Naughtiness Rating Calculated for Idiots

6. Goofy Graphs

 Graphs               

When graphs are pretty but provide no real information or are misleading.

Graphs frequently appear in the media with no scales, odd scales or totally misleading scales.

This first graph from the Financial Review: Smart Investor magazine ( Oct 2010) has no scale. You could just scribble a line and call it a graph ( See Above) !!!! Then add a number at the end point to make it look real!

Why use a graph? It’s an investment ad selling an investment product.

The second graph shows that the recent Global Financial Crisis was not so bad. Now look at the y-scale. It’s logarithmic.

Buyer beware of goofy graphs!!!!!

 

Some more useful graphs at Bored Panda Blog:

7. Crystal Ball Maths

Extrapolation.

Extending a graph or equation beyond defined limits.

Extrapolation is in your head. Just because you have a graph and, maybe, even a formula it doesn’t mean you know all the variables. Put simply, you cannot just extend a graph into the future. You may be right or……. Great Balls of Fire! You could be wrong!

Here is an extract of an article by Jeremy Laurance in The Independent, UK, on 27th April, 2009: 

‘At its worst, it could have a devastating global impact, greater than a terrorist attack, nuclear accident or environmental disaster. The World Health Organisation estimates that a mild pandemic could cause up to 7.5 million deaths.’

If you look at the current count as of 11th October 2011 verified World Wide deaths from Swine Flu are 14, 337. (See Swine Flu Count)

Any death is tragic but according to WHO in a normal year, flu kills 12,000 to 20,000 mainly elderly people in Britain and 250,000 around the world.

For more statistics : swine flu mathspig 

8. Group Think

Statistics

Independent events can cluster.

I’ve seen this phenomena explained on the TV show Numb3rs. Fire a shotgun at a barn wall, draw a circle around a group of bullet holes with chalk and say ‘Wow! We hit the target’. Similarly, if groups of workers at a power plant, say, get a disease, the newspapers immediately blame the workplace. They could be right, but on the other hand, they could be wrong and the real cause of the illness is missed. (See  ABC Cancer Scare example.)

I’ve included the water % chart below to show you how grouping numbers after the event – namely your birth – may NOT reflect the nature of the whole.

eg. ABC abandons cancer scare building (The Age 21st Dec 2006)’

‘The ABC building in the Brisbane suburb of Toowong will be abandoned without knowing the cause of a high incidence of breast cancer cases among women staff members.

A five month investigation by an independent review panel was conducted after 12 women were diagnosed with breast cancer in the past 11 years.’

‘The study showed women who worked at the office reported breast cancer at a rate 11 times higher than the general working community.’

Catalyst ABC TV 11th Oct 2007

Cancer Clusters

‘The site was also tested for carcinogenic pollution and contamination of water supply and air conditioning. Quite simply, no scientific explanation for the cancers was found.’

Radiation levels were tested. Other possible breast cancer causes sited were shift work and the Mouse Mammary Virus.

No causal link has yet been found.

9. We’re All Gonna Die!!!!!!!

Statistics

Statistics apply to Groups not Individuals

or Why Death Rates are Over-Rated!!!

We’re All Gonna die!!!  Well, um, this statement is true. The morbidity rate for humans is still 100%. Hopefully, you will be happy, healthy and 110 years old before such statistics begin to worry you. 

Newspapers often run statistics stating that your chance of being killed by a shark is 1 in 271,257 as shown. It’s NOT.

Firstly, you know this statistic is a bit dodgy. If you do not go into the water ever you will never be eaten by a shark. You chance of being shark bait is a BIG 0.

Secondly, these statistics apply to groups not to the individual. In every 271,257 Australians 1 MIGHT be eaten by a shark. In 251, 257 surfers the numbers eaten by a shark would be MUCH greater. 

Finally, these numbers still seem high. What?  77 people are killed by sharks in Australia annually. NO. These are statistics over a lifetime. 80 years, say. Less than 1 Aussie a year is killed by sharks on average but just because one surfer has died you cannot say ‘Great! The sharks have reached their annual quota I can go back in the water.’

Statistics can NEVER tell you WHAT might happen to you!

It’s a GROUP THING. All you can learn from these statistics are the  exit-the-earth probabilities for Australians that, mostly, we don’t get to choose.

10. Dumb Media Maths

Statistics

When the numbers are totally fake but no one checks!!!!

The biggest mistake made by journalists in the media is NOT QUESTIONING the numbers or validating the source. Rubbish statistics work their way into the media and become the gospel according to everyone. Sometimes, media maths seems to be written by dummies for dummies.

In an excellent article Numbers UP: Truth About Statistics ( The Independent, UK, 9th April, 2008) Simon Usborne states ‘ Flicking through a day’s newspapers often feels like tackling a numerical assault course.’ He quotes some alarming headlines form the previous day including “Ninety-six per cent of children in European orphanages are not orphans”. “In the UK we throw away 4.4 million apples a year”.

My favourite, however, is :

“Falling coconuts kill 150 people a year” 

In 2002, in an article about the uprooting of coconut trees by lawsuit-wary Australian officials, the Daily Telegraph reported: “Coconuts… kill about 150 people worldwide each year, making them more dangerous than sharks.”

No source has yet been found confirming this statistic.

Similarly,

Researchers at the University of Pennsylvania decided to search for the source of the statistic that insisted you should drink 8 glasses of water a day. Their conclusion: “It is unclear where this recommendation came from.” In other words, they could not find any study to support the “eight glasses” claim.

So when you are drinking those 8 glasses of water a day you better look up incase you’re hit on the head by a coconut!!!!!!!