Crooked.

Lawrence Solomon: 97% cooked stats

First published in the Financial Post, Jan 3, 2011
(h/t Gary Kerkin)

by Lawrence Solomon

How do we know there’s a scientific consensus on climate change? Pundits and the press tell us so. And how do the pundits and the press know? Until recently, they typically pointed to the number 2,500 — that’s the number of scientists associated with the United Nations Intergovernmental Panel on Climate Change. Those 2,500, the pundits and the press believed, had endorsed the IPCC position.

To their embarrassment, most of the pundits and press discovered they were mistaken — those 2,500 scientists hadn’t endorsed the IPCC’s conclusions, they had merely reviewed some part or other of the IPCC’s mammoth studies. To add to their embarrassment, many of those reviewers from within the IPCC establishment actually disagreed with the IPCC’s conclusions, sometimes vehemently.

The upshot? The punditry looked for and found an alternative number to tout: “97% of the world’s climate scientists” accept the consensus, articles in the Washington Post, the U.K.’s Guardian, CNN and other news outlets now claim, along with some two million postings in the blogosphere.

Sample of 77 ‘scientists’

This is an adopted article.

This number will prove a new embarrassment to the pundits and press who use it. The number stems from a 2008 master’s thesis by student Maggie Kendall Zimmerman at the University of Illinois, under the guidance of Peter Doran, an associate professor of Earth and environmental sciences. The two researchers obtained their results by conducting a survey of 10,257 Earth scientists. The survey results must have deeply disappointed the researchers — in the end, they chose to highlight the views of a subgroup of just 77 scientists, 75 of whom thought humans contributed to climate change. The ratio 75/77 produces the 97% figure that pundits now tout.

First exclude the most suitable

The two researchers started by altogether excluding from their survey the thousands of scientists most likely to think that the Sun, or planetary movements, might have something to do with climate on Earth — out were the solar scientists, space scientists, cosmologists, physicists, astronomers and meteorologists. That left the 10,257 scientists in such disciplines as geology, geography, oceanography, engineering, paleontology and geochemistry who were somehow deemed more worthy of being included in the consensus. The two researchers also decided scientific accomplishment should not be a factor in who could answer — those surveyed were determined by their place of employment (an academic or a governmental institution). Neither was academic qualification a factor — about 1,000 of those surveyed did not have a PhD, some didn’t even have a master’s diploma.

Answer just two questions – one of them meaningless

To encourage a high participation among these remaining disciplines, the two researchers decided on a quickie survey that would take less than two minutes to complete, and would be done online, saving the respondents the hassle of mailing a reply. Nevertheless, most didn’t consider the quickie survey worthy of response — just 3,146, or 30.7%, answered the two key questions on the survey:

1. When compared with pre-1800s levels, do you think that mean global temperatures have generally risen, fallen, or remained relatively constant?
2. Do you think human activity is a significant contributing factor in changing mean global temperatures?

The questions posed to the Earth scientists were actually non-questions. From my discussions with literally hundreds of skeptical scientists over the past few years, I know of none who claims the planet hasn’t warmed since the 1700s, and almost none who think humans haven’t contributed in some way to the recent warming — quite apart from carbon dioxide emissions, few would doubt that the creation of cities and the clearing of forests for agricultural lands have affected the climate. When pressed for a figure, global warming skeptics might say humans are responsible for 10% or 15% of the warming; some skeptics place the upper bound of man’s contribution at 35%. The skeptics only deny that humans played a dominant role in Earth’s warming.

Surprisingly, just 90% of the Earth scientists who responded to the first question believed that temperatures had risen — I would have expected a figure closer to 100%, since Earth was in the Little Ice Age in the centuries immediately preceding 1800. But perhaps some of the responders interpreted the question to include the past 1,000 years, when Earth was in the Medieval Warm Period, generally thought to be warmer than today.

Definition of significant

As for the second question, 82% of the Earth scientists replied that human activity had significantly contributed to the warming. Here the vagueness of the question comes into play. Since skeptics believe human activity has been a contributing factor, their answer would have turned on whether they consider a increase of 10% or 15% or 35% to be a significant contributing factor. Some would, some wouldn’t.

In any case, the two researchers must have feared that an 82% figure would fall short of a convincing consensus — almost one in five wasn’t blaming humans for global warming — so they looked for a subset that would yield a higher percentage. They found it — almost — by excluding all the Earth scientists whose recently published peer-reviewed research wasn’t mostly in the field of climate change. This subset reduced the number of remaining scientists from over 3,000 to under 300.

Mendacious meddling with the respondent pool

But the percentage that now resulted still fell short of the researchers’ ideal, because the subset included such disciplines as meteorology, which Doran considers ill-informed on the subject. “Most members of the public think meteorologists know climate, but most of them actually study very short-term phenomenon,” he explained, in justifying why he decided to exclude them, among others. The researchers thus decided to tout responses by those Earth scientists who not only published mainly on climate but also identified themselves as climate scientists.

“They’re the ones who study and publish on climate science,” Doran explained. “So I guess the take-home message is, the more you know about the field of climate science, the more you’re likely to believe in global warming and humankind’s contribution to it.”

Once all these cuts were made, 75 out of 77 scientists of unknown qualifications were left endorsing the global warming orthodoxy. The two researchers, the master’s student and her prof, were then satisfied with the findings of her master’s thesis. Are you?

Financial Post

LawrenceSolomon@nextcity.com

Views: 72