Rick Nevin
Causation Indicator: Temporality

    At a minimum, temporality asks if the suspected cause preceded the effect, but the statistical best-fit time lags for MR, SAT scores, crime, and unwed pregnancy present an especially compelling case of temporality.  While time series comparisons can result in coincidental correlations, statistical significance changes in a striking pattern across regressions comparing MR, SAT scores, crime, and unwed pregnancy rates with different time lags for 1936-1990 blood lead trends.  MR shows no significant correlation with blood lead with lags of less than 2 or over 21 years, and statistical significance peaks at a 12-year lag, for students around 12 years old.  Statistical significance peaks at a 17-year lag for SAT scores for test-takers around 17 years old; at the peak-age of offending for violent and property crime; and at the age-bracket examined for unwed pregnancy.  These best-fit lags are all consistent with lead-induced neurobehavioral damage in the first year of life.  

    Age-specific MR trends reinforce the temporality evidence of causation.  Research in the 1980s found MR prevalence peaked in elementary and junior high school, but prevalence peaked at age 14 in 1993 and age 16 in 2006.  Declines in 6-11-year-old MR from 1996 to 2006 and age 12-17 MR from 2002 to 2006 track the same documented birth cohort decline in preschool blood lead after the late-1980s.  Shifts in the peak age of arrests shows the same evidence of causation for crime rate trends.

Website Builder