1. This study was most likely an observational study because the subjects themselves chose whether or not to drink caffeine.
 
 
 
 

2. This study is unconvincing for a number of reasons:

1) Because this is an observational study, the difference could be explained by a confounding factor. For example, doctors might have told depressed patients to quit caffeine. This would be a confounding factor because it affects both the assignment to treatment groups (caffeine vs. no caffeine) and the outcome (depressed people are more likely to commit suicide.)

2) The headline of the article strongly suggests that the results apply to all women, however, only female nurses were studied, and we don't know if this group is representative of all women. Also, we don't even know if this sample is representative of all nurses, since we don't know how it was obtained.

3) We have no idea of the suicide rates. We know the coffee group has 11 suicides, and the non-coffee group 21, but perhaps there were only 11 people in the coffee group. (This is, of course, unlikely, but without knowing the size of the groups, we don't know how to compare these numbers.) Probably, the study itself looked at the suicide rates, but this was not reported in the newspaper.