“It’s one hundred billlllllllllllllllion people….”

The country was divided into what was supposed to be 33 relatively equal clusters, each being approximately 739,000 people, or 1/33 of Iraq’s population. They then surveyed 30 households in each cluster and “assumed that every household had seven individuals” (not sure where they got that assumption, they didn’t say). However, my claim is that the 33 clusters were far from the norm of what should have been selected as a random sample. Here are some cities, populations, and cluster assignments:

Sulaymaniya – 1,100,000 – 3 clusters
Salah ad Din – 1,099,000 – 3 clusters
Karbala – 1,047,000 – 3 clusters
Missan – 685,000 – 3 clusters

I’m starting to see a pattern here. These clusters don’t seem to be approximately 739,000 people each. I have highlighted Karbala, because I have heard the name in the news alot, and it seems to be a particularly violent place. Now let’s continue.

Arbil – 1,100,000 – 0 clusters

Eh? What is unique about Arbil that would cause them to get 0 clusters? Could it be that they are a Kurdish city in northern Iraq, are US friendly, and have had very few casualties during the war? In fact, I believe that is the case. They were intentionally left out of the sample because it would significantly lower the average number of people killed. The authors of this survey instead cherry-picked the most violent places and then tried to extrapolate this data over the whole of the Iraqi population.

LINK

NOOOOOOOOOO This study isn’t biased is it??? What a frickin’ lowlife. Talk about a lack of full disclosure. – Lorica