The sociology of data

Originally published on Tue, 05/07/2013 - 07:31

I am appalled by the rejection of evidence and reasoning that erupts every day.

A few examples:

Two Harvard professors, Kenneth Rogoff and Carmen Reinhart co-wrote a study "Growth in a Time of Debt" which advanced their assertions that economic growth was likely to stagnate in a country if the ratio of government debt to GDP goes over a threshold (which they claim in 90%). Cheered by conservatives because it could be used to counter the preponderance of evidence and opinion amongst economists that austerity in current times is harmful, the study received wide coverage. It also was shown by subsequent analysis to be deeply flawed. Herndon, Ash, and Pollin at U Mass Amherst demonstrated that Rogoff-Reinhart not only cherry-picked data, but their key calculations were wrong because of an Excel spreadsheet formula error. The conclusions of Rogoff-Reinhart were not only unsupported, but using the analysis of Rogart-Reinhart demonstrated the opposite conclusion. Sadly, the true motivations of Rogoff and Reinhart (and others) then came out which amounted to "ok, well forget the data and analysis, just believe what we say."

Will North Carolina next legislate the value of π? I quote from Scott Huler (who lives in North Carolina) from his blog entry at Scientific American: North Carolina? You remember: the state against science regarding sea level rise? The state with the Department of Environment and Natural Resources head who doubts climate change science and believes oil is a renewable resource? The state that tried to appoint a head of early childhood education who believed the Fukushima earthquake might have been caused by ultrasonic waves from North Korea? That North Carolina?

The scientific method the Republican-run legislature is against now is … counting. Yep — in its desperate attempts to get rid of North Carolina’s renewable energy program, the legislature has given up the radical, liberal, lamestream, obviously subjective “science” of, um, actually counting votes. You see, when the votes were actually counted, the bill that would have removed the renewables program (and said that wind, among other things, was not renewable) died in the state house, failing to emerge from committee by an 18-13 vote. Okay, hmm … you’re Republican legislator Mike Hager, you hate the renewables program, and your bill has just been defeated by an indisputable margin of five votes. What to do … what to do? Easy. You reintroduce the bill. And when it next comes up in committee, this time in the state senate? You have a voice vote — and have your finance committee chair, Republican Bill Rabon, refuse to count the actual votes. In a voice vote so close that both sides claim they would have won if the votes had been counted, Rabon declares that the bill has passed and runs off.

No, I wish I were, but I am not making this up. We have given up counting votes in North Carolina. The Reign of Error rules supreme here.

The near complete disregard for the weight of evidence for man-made components of climate change. Wow. Need I say more?

In what has been called "the most damaging medical hoax of the last 100 years" many parents and advocates turned against the use of thiomersal (also known as thimerosal) in vaccines. Thiomersal has been used for about 80 years as a preservative in vaccines. Studies of large epidemiological datasets by Mark Geier, who has now had his medical licenses revoked, and his son, David Geier, who has been charged with and fined $10,000 for practicing medicine without a license, whipped up concerns that thiomersal may be a root cause of autism. Their work led to public outcry which continues and the unfortunate decision by many parents to not inoculate their children. These studies have been shown to be highly flawed in methodology (at best). The Geiers had a financial interest in creating a protocol based on the drug Lupron and created what has been termed a body of "junk science" in support of it.