A researcher conducted a study of 34 scientists (Grim, 2008). He reported a correlation between the amount of beer each scientist drank per year and the likelihood of that scientist publishing a scientific paper. The correlation was reported as r = −.55, p = .01.
Below are the parts that are associated with this question.
a) What does a negative correlation mean in this example? Is this relationship strong or weak?
b) What does p = .01 mean in this result?
c) What might happen to this correlation if you added one person in the sample who drank much more beer than other scientists and also published far fewer papers than other scientists?
d) Plot this association in Excel using fabricated data. Then think about whether the association might be attributable to subgroups. Replot the association, marking two different subgroups (such as male and female, or different branches of science, such as biology, psychology, and physics). In the scatterplot, is the overall association between beer consumption and scientific publication spurious, or not?
e) A popular press report about this article was headlined, "Suds seem to skew scientific success" (San Diego Union-Tribune, 2008). Is such a causal claim justified?