Namir Al-Tawil, College of Medicine / Hawler Medical University
1 June 2012
I have a comment on table 4 regarding the calculation of the percentages. When calculating the prevalence of any health problem, it must be calculated among persons of each of the categories of the studied factor. In our example, the author calculated the proportion of religion types among women experiencing violence. I think he has to calculate the proportion of violence among women of each of the religion category. For instance, out of 68 Muslim women, 35 experienced violence, so the percentage must be 35/68 X 100=51.4%, but the author put 27% while for Orthodox, 66 out of 266 = 24.8% instead of what was written (50%) which will give misleading results.
Row and column percentage calculation
1 June 2012
I have a comment on table 4 regarding the calculation of the percentages. When calculating the prevalence of any health problem, it must be calculated among persons of each of the categories of the studied factor. In our example, the author calculated the proportion of religion types among women experiencing violence. I think he has to calculate the proportion of violence among women of each of the religion category. For instance, out of 68 Muslim women, 35 experienced violence, so the percentage must be 35/68 X 100=51.4%, but the author put 27% while for Orthodox, 66 out of 266 = 24.8% instead of what was written (50%) which will give misleading results.
Competing interests
None declared