There was an error in the first sentence of the ‘Data analysis’ section on page 891, in the paper by Breslau et al. (Reference Breslau2004), for which the authors apologize. It was stated that ‘each random event was weighted by the inverse of the number of events reported’, which was not the case.
The correct version of the sentence is given below:
In addition to the standard population survey weights, described above, each random event was weighted by the number of events reported by the respondent, to adjust for variation in the selection probabilities of events due to differences in number of events across respondents.