smokerlead: Pair Differences in Blood Lead Levels in 679 Matched Pairs of...
In testtwice: Testing One Hypothesis Twice in Observational Studies

Description Usage Format Details Source References Examples

Data from the 2007-2008 US National Health and Nutrition Examination Survey. There are 679 matched pair differences in blood lead levels in ug per dl, comparing a daily smoker and a nonsmoking control. A daily smoker smoked every day for the last 30 days and smoked at least 10 cigarettes per day. Nonsmokers smoked fewer than 100 cigarettes in their lives. The data are described further in Rosenbaum (2012).

1	data("smokerlead")

A data frame with 679 observations on the following 3 variables.

lead: Smoker-minus-control pair differences in blood lead levels, ug per dl.
llead: Smoker-minus-control pair differences in logs of blood lead levels.
dose: Number of cigarettes smoked per day for the smoker in the matched pair.

Pairs were matched for age, gender, education, income and ethnicity.

The data are originally from the 2007-2008 US National Health and Nutrition Examination Survey, but the current example appeared in Rosenbaum (2012).

Rosenbaum, P. R. (2012) <doi:10.1093/biomet/ass032> Testing one hypothesis twice in observational studies. Biometrika, 99(4), 763-774.

US National Health and Nutrition Examination Survey.

data(smokerlead)
attach(smokerlead)
# Compare with Table 1 in Rosenbaum (2012).
quantile(lead,c(0,1/16,1/8,1/4,1/2,3/4,7/8,15/16,1))
quantile(dose,c(0,1/16,1/8,1/4,1/2,3/4,7/8,15/16,1))
oldpar<-par(mfrow=c(1,2))
boxplot(llead,ylab="Difference in Logs of Lead Levels",xlab="Smoker-Control")
abline(h=0,col="red")
plot(dose,llead,ylab="Difference in Logs of Lead Levels",xlab="Cigarettes Per Day")
lines(lowess(dose,llead),col="red")
detach(smokerlead)
par(oldpar)