Two big science stories have broken in recent days. One has implications for the health of millions, while the other is most likely inconsequential and wrong. Guess which attracted the bigger headlines?
The news that scientists may have detected particles travelling faster than the speed of light swept round the world with barely less alacrity. It even became the lead item on the BBC World Service's news bulletins, usually dominated by war, pestilence and death.
Around the globe, professors found themselves being asked whether Einstein really had been proved wrong, and that a scientific revolution was about to break. And pretty much all of them gave the same answers: "No" and "No".
Their scepticism is based on the fact that not even the researchers making the claim have much confidence in it. The measurements needed to reveal the speed of the particles are tricky and could easily be wrong by an amount that would mean Einstein was right all along.
The team, based at Cern, home of the famous Large Hadron Collider project, went public to try to persuade other researchers to repeat the study and find out where they've gone wrong.
We should have confirmation that the finding really is codswallop within a year. But will the debunking also make headlines around the world? Unlikely: the media rarely show much interest in the fizzle of once-dazzling news stories.
Which brings us to the case of the second big science story - the one of direct interest to millions of people. In 2009, medical researchers in America published findings in the leading journal Science that pointed to a major breakthrough in understanding Chronic Fatigue Syndrome (CFS), a mysterious debilitating disorder that affects millions of people worldwide.
The cause of CFS has been the source of long and often bitter debate. Some medics insist it is all in the mind, which infuriates many sufferers. The 2009 study made headlines because it claimed to have found a link between CFS and a virus. The researchers found that in tests on over a hundred patients with the disorder, 67 per cent carried a virus code-named XMRV, compared with just 4 per cent of healthy people.
As with the faster-than-light claim, these results were greeted with considerable scepticism. Experts asked to comment pointed to the relatively small numbers of patients tested and the risk that the results were due to contamination. Despite this, the research was hailed by the media as a breakthrough, with one science magazine declaring it one of the top discoveries of the year.
The publicity had the wholly beneficial effect of encouraging other researchers to attempt to replications of the original findings. More than a dozen tried - and failed. Now even the original team have found flaws with their work, and have partially retracted their original paper. Their lead researcher has also been fired, and allegations of mishandling of data are being investigated.
It thus now looks likely that the sceptics were right all along. The link between CFS and the XMRV virus was simply a mirage, and the real cause lies elsewhere. Not that this has attracted anything like the publicity of the original claim. Indeed, unless you have access to academic journals, you could be forgiven for believing the virus theory is still viable.
But then, you could also be forgiven for believing that people with aggressive "Type A" personalities are more prone to heart attacks, that aluminium saucepans are linked to Alzheimer's disease, or that silicone breast implants cause a host of disorders. These and a host of similar headline-grabbing claims have been debunked over the years, but with nothing like the fanfare that greeted the original claim.
Still, the media can hardly be expected to provide running commentaries on the great scientific and medical debates of the day. For that, we must turn to the academic journals.
But here's the scary thing: they too have a habit of failing to publish "boring" negative results. The result is that even the scientific community can be hopelessly misinformed about the true status of controversial claims.
The existence of this "publication bias" has been known about for years. It's been of particular concern in medical research, where claims for supposedly effective new therapies are known to have a higher chance of being published than reports of failures - despite the importance of knowing either way.
Now a new study suggests the problem is spreading. An analysis of journals by Dr Daniele Fanelli of the University of Edinburgh has found that research with positive findings is taking up ever more of the academic literature.
According to Dr Fanelli, in the early 1990s around 70 per cent of papers reported positive results. Since then, the proportion has steadily increased, and by 2007 it stood at 86 per cent.
The trend was particularly marked in the "softer" sciences like economics and social psychology. It was strongest, however, for disciplines like clinical medicine and pharmacology - precisely those already causing most concern because of the dangers of researchers being misled about the true value of therapies.
Much of the problem lies with academic journals. Many insist they simply don't have the space to publish negative findings - though in these days of online publication, that sounds increasingly feeble.
Some journals even have a policy of refusing to publish results that confirm or refute previous studies already published elsewhere.
There is also growing concern that, sensing the futility of submitting them, researchers themselves are not bothering to write up negative findings.
When - as seems very likely - the Cern team's faster-than-light claim is debunked, most newspaper editors will give the story short shrift. But if researchers increasingly follow suit with their own negative findings, it will not just leave us ill-informed. It will undermine the scientific process, the best means we have of distinguishing fact from fiction.
Robert Matthews is Visiting Reader in Science at Aston University, Birmingham, England