Tuesday, October 2, 2018

Science: Understanding Statistics


Numbers are meaningless by themselves for people that are not studying arithmetic or a related subject.  But we can all compare numbers and determine if one is bigger or smaller.  There is an old caution about people using numbers that notes "figure don't lie, but liars can figure."  Dr. Clark, teaching a biostatistics class, and later a member of my doctoral committee was comfortable telling an old joke about the statistician that drowned crossing a stream that averaged one foot deep.  I am currently puzzled by how I would know how many beers before I would sleep or blackout.  I haven't had a beer in over thirty years, and I found out over thirty years before that, that four was too many for me in one evening.  I seldom drink now, just an occasional small glass of wine or a drink traditionally garnished with olives, although I much prefer a glass of water.

Divorce rates

Our local paper has a front page story today, October 2, 2018, having the largest and first title "Millennials are causing the U.S. divorce rate to plummet".  The author of the article credits an analysis by a sociology professor using "U. S. Census Bureau survey data.  Others are cited who note boomers had married earlier and had a higher divorce rate.  Other factors are discussed that show how commitment and employment opportunities for women may be reflected in the numbers.

Toxic substances

Newspapers, magazines, and news programs are full of frightful stories about contamination or dangers of our food, water, medicines, and environment.  You are right to be concerned, and something should be done, but realize presence of some dangerous substances can be determined to number of parts per trillion (ppt).  Back fifty+ years ago, when I was a member of the Midwest Benthological Society and a graduate student studying bottom dwelling aquatic organisms and water pollution, few substances could be identified at concentrations less than one part per million,

Exposure limits

When tests associate the lowest level at which damage to test organisms is shown, permissible levels were typically set by regulatory agencies at one percent of the largest exposure which produced problems.  Consideration should be given to other factors such as is it accumulated in tissues of an organism, is it persistent or does it break down naturally, how it interacts with other substances, and is it degraded or converted into other substances of concern.

 An extreme example

In the early 1960's when I was teaching at Western, I was prepared to testify at hearings for the planned Palisades Nuclear Power plant regarding their proposal to release very slight concentrations of radioactive elements in the expected cooling water releases into Lake Michigan.  Zero release was part of the final plan.  The seemingly small amounts had some elements that previous research (by others) had been shown to be concentrated by 30,000 times over the concentration in which they lived.  Take that concentration and note how they might be fed upon and concentrated in the food chain, or get filtered out an the beach, and the possibilities are frightening.  Fortunately, no releases were permitted.

In most cases, one part per trillion (a million of them to make one part per million) is not of itself a big danger.  But it does mean it might be coming from a more dangerous source that should be identified.

Food labels

Helpful, but ingredient lists can be more misleading than percent daily values.  Your jam or jelly may list a fruit first showing it is the most abundant food in the product.  But when you add up the later listed sugar, fructose, glucose, honey, and water, it would be possible that less than 20% is fruit and over half could be some form of sugar.  And if you are diabetic, you should know something about the sugar content of fruit.

Joseph Engemann,  Emeritus Professor of Biology, Western Michigan University, Kalamazoo, Michigan    October 2, 2018

No comments:

Post a Comment