Computerized Superstition

SECONDARY CORRELATION

aka COMPUTERIZED SUPERSTITION

see many studies in newspapers and on TV reported as fact that I can't believe to be true. Many of these studies are the result of computer correlation studies in which essential data is missing. They show a cause and effect relationship between events that are related only by being effects of an unidentified cause.

We are born with the ability to correlate events. We learn as a child that if we touch a hot iron, we will get burned. In other words, there is a direct cause and effect correlation between the action and the result. Sometimes we correlate unrelated events such as seeing a black cat and having bad luck just because a correlation just happens to occur. It is also statistically possible to win at a race track after picking a four leaf clover, though not probable.

With the great proliferation of computers, there is big business in correlation studies. Many of these studies correctly detect direct cause and effect correlations between hazards and injury such as that between drinking and driving. What concerns me is the way computers are used to develop correlations without all the facts. With insufficient data, correlations appear when in fact there is no direct cause and effect, but rather a common, but unidentified cause to two unrelated effects.

There have been may newspaper articles about the supposed correlation between Video Display Terminals (VDTs) and various diseases such as heart disease, cancer and birth defects. Many of these reports are based on computer analysis where a large number of statistics are taken of people who work at VDTs and those who do not. Suppose that the problem is that the nature of long sedentary work invites the use of stimulants such as coffee, tea, and cigarettes. The use of these stimulants is probably not included in the statistics used to correlate VDT use with various diseases. There is certainly a correlation between the use of these stimulants and many diseases. Now suppose there is also a small correlation between VDT use and stimulants. Then the computer analysis will show a correlation between VDT use and diseases even though a direct cause and effect relationship does not exist. This is secondary correlation, an indirect correlation to a common, often unreported, variable.

Another example that was in the news was a supposed correlation between low frequency electromagnetic fields (EMF) and diseases. A newspaper reporter was excited when an alternating field of 35 milligauss was found in elementary school classroom. What she didn't know is that the Earth's magnetic field is almost 500 milligauss, and that magnets used to pin notes to the refrigerator are over a million milligauss. The computer correlation of magnetic field to children's disease probably failed to include, report and correlate the use of carcinogens (creosote, fungicides, etc.) in wooden power poles. Again secondary correlation may occur because electromagnetic fields are found near power poles, and the frequent use of poisonous materials used to preserve wood power poles can cause medical problems. Has there been a study to see if there is a difference in the occurrence of cancer between linemen who work on wooden power poles versus linemen who work on metal poles, or linemen who work on buried power lines? But even working on metal poles can expose one to the herbicides and defoliants sometimes used to minimize weed growth around high tension lines. Old power transformers leaking a poly-chlorinated biphenyl (PCB) are fortunately a thing of the past, but may have influenced past EMF studies. There is even evidence that corona discharge (leakage) over insulators (used in very high voltage transmission lines) produces X-ray energy.

Another problem is determining what is cause and what is effect. Some people incorrectly assume the Carbon Dioxide causes Global Warming. But a close look at the data shows that Global Warming occurs before the increase of Carbon Dioxide. Suppose that Global Warming is caused by an increase in Solar Radiation from the Sun, and the increase heat causes the Earth to warm, which causes the oceans to warm and release some of their large store of dissolved Carbon Dioxide as well as the greenhouse gas called 'water vapor'. The result would be wetter weather and a significant increase in plant growth. With more CO2 in the air and more water, more food could be grown.

I hear people worried about the ozone hole that forms every year over the Antarctic. 'Why?' I ask. The say they are worried about ultraviolet light. It is true that ozone attenuates UVA & UVB, but what they do not realize is that UVC is needed to replenish the ozone that was naturally converted back to ordinary oxygen. It is only during the long Antarctic night that the ozone hole can form. During the night there is no UVA, UVB or UVC from the sun. When the sun reappears in October, the ozone reforms from the UVC and the hold is closed.

Over and over again I see false correlation studies reported as fact because the computer said so. I wonder what data was used in the study, and what data should have been taken, but was not. I refuse to believe wild cause and effect claims until the mechanism of the cause and effect is adequately explained.


I recommend reading chapters 5 in the book 'Kicking the Sacred Cow' by James P. Hogan 2004. [ISBN 0-7434-8828-8] What he says in this chapter is similar to what I have known for years.


Russ' main page

This page's URL.is http://Opinion.FarTooMuch.Info/supersti.htm

This Web Site was designed by Russ.