In the article, “The political ecology of lead poisoning in eastern North Carolina” (Hanchette, 2007) Hanchette studied the ecological impact of lead poisoning in eastern North Carolina and childhood blood lead levels.
The elevated blood lead levels dropped from 88% to 2.2% from the late 1970s to 2000. But however, this is not the case in many eastern North Carolina counties.
Hanchette used LISA and Moran's I statistical analysis to spatially interpret the data. The data showed that the high childhood blood lead levels were concentrated in the eastern North Carolina counties while many of the Appalachian counties had very low childhood blood lead levels. Hanchette compares these distributions with variables such as poverty levels and age of houses. Those who live in poverty may not be able to receive adequate medical treatment and suffer from the effects of high childhood blood lead levels. And those who live in older houses are at risk of having water pipes made out of lead, which could also affect the childhood blood lead levels. Hanchette finds that the eastern counties affected with high childhood blood lead levels correlate with the growing of tobacco in the region. For decades, tobacco has been growing in eastern North Carolina, and Hanchette suggests that this may be the cause of the high childhood blood lead levels.
Correlation does not imply causation. As the distribution of African Americans in North Carolina correlate with the counties affected by high childhood blood lead levels. Hanchette justifies this by providing historical evidence of tenant farming and segregation in these counties.
I believe Hanchette does a great job with the statistical analysis, and takes many variables into account (such as the poverty rate and age of housing). And this paper does an excellent job of showing us why high childhood blood levels are dangerous. However, this paper does have its flaws. The first paragraph from the paper has almost exactly the same wording the abstract. I believe that this should be changed, and the abstract should provide more results instead of more background information. Also, I believe there should be more graphs depicting the distributions of other variables, perhaps showing change over time?