http://gu.com/p/3t99n
Scientists who study human behaviour are more likely than average to report exaggerated or eye-catching results if they are based in the United States, according to an analysis of more than 1,000 research papers in psychiatry and genetics. This bias could be due to the research culture in the US, authors of the analysis said, which tends to preferentially reward scientists for the novelty and immediate impact of a piece of work over the quality or its long-term contribution to the field.
Showing posts with label statistics. Show all posts
Showing posts with label statistics. Show all posts
Tuesday, August 27, 2013
Wednesday, July 24, 2013
Wednesday, September 12, 2012
Can tablet screens disrupt sleep? (and how to report it)
Can tablet screens disrupt sleep? | Technology | guardian.co.uk
what is worth reading about this 'science story' is not the interesting headline (in the classic 'study shows' genre) but how the piece itself is an unusually good example of how such stories should be reported. While there is indeed first the eye catching headline, the inevitable invoking of authority with phrases such as 'a team of scientists', and even mention of a related proven phenomenon which makes the study more plausible ("Melatonin is a hormone used by the body as a biological indicator of how dark it is outside"), what is excellent about this piece is it then provides details on the study (just 13 volunteers), links to the paper itself, who funded it, and most importantly, a counter argument, which in fact is much more persuasive than the original conclusion.
This story is a classic example of small, unreliable and commercially funded 'research' which we are so often bombarded with, not because they we need to be informed, but because they are 'topical', and help fill some pages. So refreshing to see it handled properly for a change.
Some extracts
http://www.telegraph.co.uk/health/healthnews/9533249/Using-iPads-before-bed-can-lead-to-a-poor-nights-sleep.html
With its classic 'researchers are warning approach" - though fair dues to the Telegraph for squeezing in a cancer reference, since as it so often tells us, everything must either cause or cure it...
what is worth reading about this 'science story' is not the interesting headline (in the classic 'study shows' genre) but how the piece itself is an unusually good example of how such stories should be reported. While there is indeed first the eye catching headline, the inevitable invoking of authority with phrases such as 'a team of scientists', and even mention of a related proven phenomenon which makes the study more plausible ("Melatonin is a hormone used by the body as a biological indicator of how dark it is outside"), what is excellent about this piece is it then provides details on the study (just 13 volunteers), links to the paper itself, who funded it, and most importantly, a counter argument, which in fact is much more persuasive than the original conclusion.
This story is a classic example of small, unreliable and commercially funded 'research' which we are so often bombarded with, not because they we need to be informed, but because they are 'topical', and help fill some pages. So refreshing to see it handled properly for a change.
Some extracts
- - team of scientists who found that exposure to the light from self-luminous displays, such as Apple iPads, could be "linked to increased risk for sleep disorders because these devices emit optical radiation at short wavelengths, close to the peak sensitivity of melatonin suppression".
- - However, Russell Foster, a professor of circadian neuroscience at the University of Oxford, is not convinced. For a start, he said, melatonin levels were not a good way to predict how much sleep someone will get
- - One final thing that very few of the news reports on the research have mentioned: the study was funded by Sharp Laboratories of America, the research organisation run by Sharp Corporation, the electronics company that makes TVs, PCs and tablets.
http://www.telegraph.co.uk/health/healthnews/9533249/Using-iPads-before-bed-can-lead-to-a-poor-nights-sleep.html
With its classic 'researchers are warning approach" - though fair dues to the Telegraph for squeezing in a cancer reference, since as it so often tells us, everything must either cause or cure it...
Monday, April 2, 2012
the lotteries in life
Always think articles on statistics and chances worth reading - since we all seem to have a mental blindspot when it comes to probability
http://www.guardian.co.uk/commentisfree/cifamerica/2012/mar/30/how-mega-millions-is-like-shark-attack
Though of course some odds aren't easily compared; even if chance of a shark attack for the global population is 250m to 1, the odds must be significantly lower for people inland who never enter the water! But still, does make the important point that we (a) are terrible at judging odds in the first place and (b) allow 'ease of visualization' (influenced by media reports) to distort our judgement even more.
http://www.guardian.co.uk/commentisfree/cifamerica/2012/mar/30/how-mega-millions-is-like-shark-attack
Though of course some odds aren't easily compared; even if chance of a shark attack for the global population is 250m to 1, the odds must be significantly lower for people inland who never enter the water! But still, does make the important point that we (a) are terrible at judging odds in the first place and (b) allow 'ease of visualization' (influenced by media reports) to distort our judgement even more.
Subscribe to:
Posts (Atom)