Hmm I have a feeling there's a miscommunication here somewhere.
I think this might be one instance where you explained why you didn't agree with my dismissal of the response rate:
Because it compares a war-ravaged town in an impoverished country which has been at war on and off with it's neighbors and itself for 20 years or so with three peaceful and significantly wealthier countries. This was pointed out to you before.
Yes, this could be true. But the
specific issue with the response rate is that a low response rate could bias the survey's results if non-respondents have characteristics in common with each other that set them apart from respondents.
In the case of the report's finding that Fallujah had a high death rate for babies, non-response would affect this finding if non-respondents had a lower death rate than respondents.
Using the hypothetical possibility that every non-respondent surveyed had 0 baby deaths, and incorporating this into the survey's results, it seems that Fallujah
still has a higher death rate than other places.
Yes, this could hypothetically be because Fallujahns smoke more.
Yes, this could hypothetically be because Fallujahns are all lying about baby deaths just to make America look bad.
Yes, this could hypothetically be because Fallujah is full of insurgents who were making chemical bombs when the US liberated the city.
Yes, this could hypothetically be because wizards cursed the city.
BUT
It means that the report's findings of higher rates of baby death in Fallujah are
not invalidated due to the response rate to the survey,
unless I did something wrong with the math.