Sexed up numbers not to be trusted

Vladd67 said:
‘Sexed-up’ numbers should not always be accepted as science - Times Online
Lets face it everyone twists statistics to fit their argument.
Click to expand...

No, not everyone does. As the article itself points out, the danger isn't in the numbers, or in using the statistics garnered from scientific studies, but in the way these are often -- but by no means always -- abused. If everyone did it, we couldn't trust the statistics at all, and such is certainly not the case. But we should take a leaf from the page of science, and use critical thinking and a dash of skepticism, to ascertain how accurate the reporting of scientific information may be....
 
I find it funny that media is reporting that we should not trust such numbers when they are one of the biggest groups responcible for doing so - just after governments ;)
Its painful, but true that so much progress is damaged by media hype and number altering - I think one paper was stating this year that the Ice masses in the Arctic were growing - - they had only used the data for a single shot period which showed increased precipitation leading to icemass growth - ignoring the following summer and the melt that would occure!
 
Lies, d**d lies and statistics...

Um, we were taught that after using Excel's wondrous curve matching facilities to match our data to our hypothesis, we turn the problem upon head and attempt to match the opposite notion...

This lack is why there's a recurring scandal over clinical trials that don't really prove anything beyond that researchers ran them...
 
We were always warned to try to fit the data to the contrary hypotheses, too.

And we had the dire experience of graphing a most curious set of data-- Sure there was a scatter, but log/log seemed to give least residuals. Imagine our chagrin when told those numbers were actually random...
 
A third of road fatalities in the UK are caused by drunken drivers. Therefore two thirds are caused by sober drivers. The solution is obvious ...
 
...get all the drunk cyclists and drunk pedestrians off the road!
 
This thread title is so misleading...here I thought it might be an interesting discussion. Then comes statistics!

Joking aside, despite the admitted and oft-discussed flaws of statistics, a thorough and unbiased statistical analysis of a situation with vastly more information than a human mind can process unaided can be a wonderful thing. Without advanced statistics the Human Genome Project would never have been completed...at least not for several more decades.
 
The moment people start quoting statisitics is the moment to reach for you handy copy of How to Lie with Statistics (Darrell Huff) to find a dozen reasons not to believe said statistics :)
 
You must log in or register to reply here.
Thread starter Similar threads Forum Replies Date
Oligonicella The importance of numbers SFF Lounge 0
M Interesting thought experiment on impact on global weather of vast numbers of solar panels Science & Nature 13
psychotick It's A Numbers Game! Publishing 0
Elckerlyc The 100 Word Anonymous Challenge Roll of Honour - Facts & Numbers Workshop 10
M Why animals can recognise numbers but only humans can do maths. Science & Nature 0

Similar threads


Back
Top