The latest United Nations Relief and Works Agency (UNRWA) report on infant mortality in Gaza deliberately misrepresents a scientific study on the subject, the Israellycool blog revealed Sunday – then fed that report to several international news agencies.
UNRWA sent a press release with the study to major news agencies claiming that infant mortality rates in Gaza are on the rise; several of the agencies published the release immediately.
However, as Israellycool notes, there are several gaping flaws in the study:
- Whereas the UNRWA indicates that the study examined facts on the ground in Gaza on the large scale, the scientific paper itself states that it only spoke to a very specific subset: mothers with two living children who reported to UNRWA health centers.
- The number of subjects interviewed is so low that the paper itself admits at the end that while a minor increase was seen in infant mortality rate, "the infant mortality rate could in fact be stable or continuing to decline."
- Moreover, as the blog staff points out, the subject pool is so low that the p-value – or the probability that the exact same results would appear by random chance – is a whopping 0.61, or 61%. Statisticians reportedly dismiss any study with a p-value below 0.05 (5%) as too weak to be reliable.
- The journal in which the study was published is an open-access online journal which apparently accepts all papers offered and then peer-reviews them – a practice which would usually deem such a journal as a source too unreliable to use within any respectable academic context.
Arutz Sheva reached out to UNRWA spokesman Chris Gunness over the findings, but has so far not received a response.