27th July 2012
Observer bias is the well-known tendency for a researcher to draw conclusions from a set of evidence based on his or her own preconceptions or understanding of the problem; a clear manifestation of the inevitable trumping of the idealised objectivity of science by human nature. A well-known example of this is in the measurement of the charge carried by an individual electron, one of the basic constants in physics.
In 1913, the American physicist Robert A Millikan published his estimate of the quantum of electronic charge, measured in his seminal ‘oil drop’ experiment. The importance of this was recognised by the award of the Nobel prize for physics a decade later. But the currently accepted value is slightly higher, only by less than 1% but nevertheless by an amount significantly greater than the standard error for Millikan’s experiment.
The change in the accepted value is because Millikan used an incorrect value for the viscosity of air in his calculations. Richard Feynman, often thought of as one of the best physicists who never won a Nobel prize, talked about what happened next at a commencement address at Caltech in 1974: “It's interesting to look at the history of measurements of the charge of an electron, after Millikan. If you plot them as a function of time, you find that one is a little bit bigger than Millikan's, and the next one's a little bit bigger than that, and the next one's a little bit bigger than that, until finally they settle down to a number which is higher. Why didn't they discover the new number was higher right away? It's a thing that scientists are ashamed of - this history - because it's apparent that people did things like this: When they got a number that was too high above Millikan's, they thought something must be wrong - and they would look for and find a reason why something might be wrong. When they got a number close to Millikan's value they didn't look so hard.”
Such things can and do happen all the time. However careful scientists are, there is a tendency to edit their own results, removing outliers from the data and therefore sometimes ignoring things which later turn out to be significant. Scientists are human, and they make judgements. This tendency is even more marked when they look at other people’s work. Understandably, they often dig their heels in when data appear to conflict with their own understanding of things, while unquestioningly accepting new supportive evidence. At present, nowhere is this truer than in the contentious world of climate change.
Take, for example, a news release from NASA this week, Satellites see unprecedented Greenland ice sheet surface melt
, reported by the BBC and others. It seems that a number of satellite observations confirmed that 97% of the total area of the Greenland ice sheet had shown signs of melting on July 12, the highest seen in three decades of remote observation. Normally, only about half of the total shows surface melting in the summer. Most of this is, as expected, at lower altitudes, but this year there was even some melting observed at the Summit station, about two miles above sea level.
As an observation, this is both interesting and worth bringing to people’s attention, but the inescapable subtext is that this is highly unusual and likely to be a sign of human-driven climate change. Only towards the end of the news release do we find that the temperature at the top of the icecap “hovered above or within a degree of freezing for several hours July 11-12”
, so we have to assume that much of the melting was extremely superficial. We also learn that the ice cores obtained from the summit have revealed that such melting has occurred roughly once in 150 years in the past, with the last event in 1889.
The cause (at least this time) seems to be an unusually strong ridge of warm air. Putting this in context, Greenland has suffered extreme weather at the same time as the heatwave across much of the USA and the record wet summer in northern Europe. Nevertheless, to reinforce the implicit message, one of the NASA researchers is quoted as saying "But if we continue to observe melting events like this in upcoming years, it will be worrisome."
During the last week, we have seen two other related stories from the BBC: Iceberg breaks off from Greenland’s Petermann Glacier
and Antarctic: Grand Canyon-sized rift ‘speeding ice melt’
. In the first case, an iceberg ‘twice the size of Manhattan’ has calved, following another twice the size two years ago. The second story reports a geological feature which allows incursion of seawater to speed up melting of part of the west Antarctic ice sheet. Such observations are the sorts of things scientists do, but it is only the subject which makes them newsworthy. This in turn encourages organisations to issue more press releases, a number of which are then reported on by journalists in a self-supporting cycle (or perhaps an example of positive feedback?).
Contrast this, then, with another story, Did exploding stars help life on Earth to thrive?
, a news release about a new publication in the Monthly Notices of the Royal Astronomical Society by the Danish astronomer Henrik Svensmark (Evidence of nearby supernovae affecting life of Earth
). By reconstructing the intensity of supernovae over the last 500 million years and comparing this to sea level, atmospheric carbon dioxide level and the fossil record, Prof Svensmark came to the conclusion that periods of relatively high supernovae activity have been good for biodiversity, probably because the relatively cool conditions created a greater variety of habitats across the latitudes. However, particularly intense activity appears to have moved the Earth into the Ice Ages. Supernovae activity and sea level are together enough to explain biodiversity changes across geological time.
This hypothesis is, at least on the surface, far more important to our understanding of the evolution of climate than recent observations of polar ice, but has received little if any attention in mainstream media. The reason seems to be that this paper gives additional credence to Svensmark’s hypothesis that cosmic rays (originating from supernovae) have a major influence on cloud formation and hence overall temperatures and weather patterns. Supporters of the mainstream IPCC view that carbon dioxide is the dominant influence have been quick to dismiss this alternative hypothesis (which is clearly explained in Nigel Calder’s book The Chilling Stars
Now, Svensmark’s ideas may turn out to be wrong, but supporting evidence has been mounting (including from the CLOUD
experiment at CERN) and, to any objective observer, this approach remains a credible scientific hypothesis, worthy of further work. The inescapable conclusion is of observer bias in those who are so keen to ignore it. This is bad for science and, ultimately, for society; wilfully dismissing evidence which conflicts with received wisdom results in misguided policymaking and taxpayers’ money is wasted on futile projects. A little more objectivity would be good for all of us.
This will be the last regular Scientific Alliance newsletter until the end of August.