Lisa Jo Rudy was absolutely right to compare the logic of Monty Python and the Holy Grail to a recent study by two Cornell business professors asking Does Television Cause Autism? A judge named Bedevere and some villagers try to convince themselves a young woman is a witch because witches and wood float as do ducks. If she weighs as much as a duck, she must be made of wood, which proves she is indeed a comely hag. After rigging the scales, the witch and duck balance perfectly, and the villagers haul her away to the stake. Bedevere says to King Arthur, the smartest of the bunch, "Who are you who are so wise in the ways of science?"
Before getting to the study, let me make you wise in the ways of statistics (I have a master’s degree in that field). If two variables have a correlation, one does not necessarily cause the other. A professor at a small college in Pennsylvania requires his students to analyze statistics on life expectancy from The World Almanac and Book of Facts 1993. The variables in this table include life expectancy at birth, the number of television sets per person, and the number of physicians per person. When calculated correctly, the statistics correlate the number of television sets per person and life expectancy. Aha! TV deprivation may cause early death and, to solve this problem, the United Nations ought to ship televisions into these deprived nations. Right? Wrong! In this case, an underlying factor (poverty) is driving both variables (television and life expectancy) making them appear to be related.
The professors studied many variables related to climate and television viewing by using a Bureau of Labor Statistics’ American Time Use Time Survey: precipitation, hours of daylight, ethnicity, military service, income, education, etc. Did you notice anything missing from this list? That is right! The actual survey did not include autism. Statistical number crunching correlates precipitation and television viewing.
Buried in Table 2 of the study and left out of the most media coverage of this issue are the other factors found to be statistically significantly correlated (the ones with three stars ***): hours of daylight, bachelor’s degree, and employment. Why does television trump these other variables? People who live in rainy places or have less daylight watch more television. People who work or have bachelor’s degrees watch less television. Imagine how foolish the headlines would look: “Does darkness cause autism?” “Can employment prevent autism?” “Can college prevent autism?”
Stay tuned for the next installment of Tammy Glaser's special coverage of the link between autism and television . . .