6. Correlation, causation, and consequences


Preamble: Some commentators on mathematical thinking claim that probability and statistics are more important to teach than calculus (e.g., Arthur Benjamin). This activity adds critical thinking to the very outset of getting students to think with statistical concepts. (Notice that, unlike many accounts of statistics, there is no reference here to probability at the foundations. Instead, we start with observations and patterns in them.)

Preparation
Print out this page of heights data
Read the precis below of Chase, "False Correlations = Real Deaths."

Goals

Activity
1. Interactive Mini-lecture: visual aids, including interpreting parent-offspring height patterns
On social embeddedness of quantitative reasoning:
  • The causes proposed reflect social actions desired/ supported
    • because
    * the categories used (& thus data collected)
    * assumptions needed to make causal claims from patterns
      • are places where social assumptions can influence the scientist's decision, &
      • science cannot be done without decisions in these areas.
  • When people make explanations look at how they attempted to identify or locate the causation and consider alternative styles of causal explanation. If a scientist is emphasizing, say, a unitary cause, consider the multiple factors they are excluding.
    This approach will help you raise questions about the other commitments that influence their choice of questions, categories, factors, and admissable explanations. The causal explanation that is advanced often corresponds to the person's commitments to certain forms of social action, e.g. Galton (Darwin's cousin) didn't measure any environmental variables and was thus able only to reach conclusions about (supposedly) inborn characters; Davenport and others similarly denied the significance of Goldberger's experimental evidence for dietary basis of pellagra.

  • Identification of more proximate causes (e.g., a lung cancer gene) is not necessarily needed to ensure the most effective action (e.g., we can encourage people to stop smoking independently of knowing the genetic mechanisms of lung cancer).
2. Case: What causes a disease?--the consequences of hereditarianism in the case of pellagra; discussion and analysis

3. Instructor-led discussion of different styles of causal explanation and their relation to ideas about politics and social action. (Relevant schema about styles of causal claims.)
4. Plus-delta feedback on activity

PRECIS (full chapter in password-protected readings)
Chase, Allen. (1976). “A Few False Correlations = A Few Million Real Deaths: Scientific Racism Prevails Over Scientific Truth.” The Legacy of Malthus. The social costs of the new scientific racism. New York, Knopf.

Misdiagnosed for centuries as leprosy, scurvy, syphilis, and other diseases marked by rough red skin eruptions, pellagra was described by the Spanish physician Gaspar Casal in 1735 as a distinct disease that he called mal de la rosa (the red disease). It was, in each country, a disease of the very poor, in which the skin eruptions were followed by debilitating sequences of diarrhea, lassitude, dizziness, and various mental disorders ranging from depression to violent lunacy.

In 1906, George H. Searcy, a southern physician, discovered many cases of pellagra among the inmates of the Mount Vernon insane asylum in Alabama. Doctors and public health officials started to seek out non-institutionalized cases of the red disease and soon realized that pellagra was endemic throughout the poorest section of the nation, the southern states. By 1908, various states had established pellagra commissions, and a new National Association for the Study of Pellaga held the first of its annual scientific meetings devoted to reports and discussions on the causes and management of the new plague.

A major treatise by Dr. Edward Jenner Wood in 1912 revealed that “the cause of pellagra is still unknown…” although there abounded a number of hypotheses at that time. Throughout medical literature, however, there was one constant thread of evidence suggesting that pellagra was caused by the inadequate diets of poverty and that in the presence of enough food, the disease never struck. Although he reviewed many of the early European authors who had written about pellagra and diet, Dr. Wood remained resistant and wrote that, “I can deny that bad nourishment in the sense of insufficient nourishment is a cause.”

In 1912, a Pellagra Commission was established under the auspices of the New York Post-Graduate Medical School and Hospital, initially under the direction of Joseph F. Siler, a physician with the Army Medical Corps, and Philip E. Garrison, a Navy surgeon. The following year, Dr. Charles B. Davenport (a zoologist by training) from the Eugenics Records Office in Cold Springs Harbor, New York, joined forces with the Commission with the intention to study pellagra from the viewpoint of heredity as a causative factor. Around the same time, due to the continuing social and economic problems caused by pellagra (disease-caused absenteeism in the textile mills; the loss of European markets for American corn due to the mistaken belief that corn was responsible for the disease), the Surgeon General of the United States appointed bacteriologist Joseph Goldberger of the U.S. Public Health Service to discover the cause and cure for pellagra. Coming from an impoverished immigrant family background himself, Goldberger had a greater insight into the total environment of poverty than did Siler or Garrison. He, too, set up his fieldwork in Spartanburg, South Carolina, where Siler and Garrison were working, but he took the time to travel around the district, visiting the insane asylums, the orphan asylums, mill towns and rural slums where pellagra was rampant. He noticed two things: that whenever there was great poverty, there was pellagra, and that the better off employees of the prisons, asylums, or mills never developed a single case of pellagra. Although the disease had been known about for far longer than other diseases, pellagra was possibly never considered within the ranks of the great plagues such as malaria, yellow fever, or typhoid because it struck only the poorest.

Goldberger set out to prove his hypothesis that pellagra was not a communicable disease, but was instead caused by the lack of an unknown vitamin which he called the PP (Pellagra Preventative) factor – as well as meats, poultry, fish, dairy products, fruits and vegetables in the diet. He conducted a number of famous experiments in Mississippi with volunteer white prison inmates in whom he first induced pellagra by withholding and then cured them by re-introducing meats, vegetable, and eggs into their diet. In another dramatic experiment, Goldberger, his wife Mary and his assistant Dr. G.A. Wheeler injected themselves with biological materials from pellagrins consisting of blood, vials of skin scrapings and ingested little dough balls impregnated with the urine and feces of the patients.

The publication of Goldberger's report on his Mississippi experiments came as something of a blow to the Pellagra Commission in which Siler, Garrison, and Dr. Ward T. MacNeal were working. At the same meeting where Goldberger and his colleagues reported their findings, MacNeal offered his report, which identified lack of plumbing and sewerage as the culprit in pellagra. Goldberger pointed out that the relationship of pellagra to sewerage might be a case of “false correlation” since the well-to-do [and better fed] had the sewers in the town and the poorest [undernourished] citizens lacked plumbing and sewerage. His argument was dismissed by MacNeal who proposed that the analysis of mass data by statistical correlations – a method adopted by Davenport – was preferable to just “theorizing” about the data.

Despite the convincing case made by Goldberger for a diet-related cause to pellagra, the mortality figures for the disease continued to rise in the years following his discovery. Among the several reasons for this disappointing result, one stands out: the efforts by Charles B. Davenport to correlate the incidence of mental deficiency and insanity associated with pellagra to his view that pellagra was an inheritable disease. Davenport's thesis was that pellagra was “the reaction of the individual to the poisons elaborated in the body, probably by a parasitic organism.” To reach this conclusion, he also had to explain that although pellagra was not an inheritable disease in the sense that brown eye color was inheritable, the course of the disease did indeed depend on certain “constitutional, inheritable” traits of the affected individual. “Susceptible” parents would produce children who were also susceptible to the disease.

Davenport and Elizabeth Muncey, M.D., published an article on the hereditary causes of pellagra in the Archives of Internal Medicine. Subsequently reprinted as a Bulletin of the Eugenics Record Office, it made a third and final appearance in the final report of the pellagra Commission. In this report, Davenport and Muncey presented their detailed tables and pedigree trees (the ‘black charts' of eugenics) that purported to establish their view that pellagra was a hereditary disease. These pedigree charts only succeeded in proving that the children and grandchildren of poor people [would] quite predictably suffer the socially preventable deficiency diseases of poverty. The size (444 pages), scientific complexity, and ‘authoritative’ tone of the Davenport report combined with the low standard of medical education that produced physicians who did not have the capacity to question the report's veracity, meant that most accepted the conclusion that it was “bad genes” and not “bad diet” that was responsible for pellagra.

Society, specifically the public health policies in society, was thus let off the hook in terms of acting to increase the living standards generally among the working poor. The excuse for doing nothing to improve the economic opportunities of the poor in the South which, in turn, would have allowed people to earn enough money to adequately feed themselves and their families – was justified by the ‘inborn inferiority’ of the poor. Ironically, it was not until the Great Depression when the government started disbursing food and welfare funds to save the lives of all people rendered hungry and helpless, that Goldberger's pellagra prevention factor was able to protect all classes of people.

The Great Pellagra Cover-Up fro 1916-1933 was the result of “scientific racism.” Had Goldberger's findings been acted upon, millions of completely avoidable premature deaths, chronic degenerative diseases and other disorders known to be associated with chronic undernourishment could have been avoided. Such was the missed opportunity that benevolent social action should have had on the clinical reactions to malnutrition.
(Precis by Jan Coe)