Correction to: Journal of Exposure Science and Environmental Epidemiology (2010) 20, 101–111; doi:10.1038/jes.2009.5

In the above manuscript, a correction needs to be made to Eq. (7):

Instead of:

Using a two-sided test, the power to detect a significant positive association based on a surrogate model given the power of the true exposure model is

… and the value 1.96 is the critical value under the normal distribution function for a two-sided test.

It should read:

Using a one-sided test, the power to detect significant positive associations based on a surrogate model given the power of the true exposure model is:

… and the value 1.96 is the critical value under the normal distribution function for a one-sided test with significance level of 2.5%.