For question 1a, it states a subsample of the participants had urinary cotinine measured to validate self-reported smoking data. However, in the table of this validation study it appears 1330 self-reported smoking and 825 self reported not smoking. However, in the complete sample listed above, only 883 self reported smoking. How did the subsample contain more self-reported smokers than the entire study sample? If this was actually a separate sample for calculating sensitivity and specificity than the study we are interested in, can we assume the values apply to the current study for QBA?
For question 2b, it appears the depression prevalence in the exposed and unexposed is reversed from question 2a as it states: "depression prevalence among the exposed and unexposed were 0.2 and 0.4, respectively" while in 2a, it states prevalence of depression is 0.4 in exposed and 0.2 in unexposed. Should we assume the prevalences provided in 2a are intended (prevalence of depression is 0.4 in exposed and 0.2 in unexposed)?
Thanks!