[lit-ideas] Re: statisticions, anyone?

  • From: "Andreas Ramos" <andreas@xxxxxxxxxxx>
  • To: <lit-ideas@xxxxxxxxxxxxx>
  • Date: Thu, 12 Oct 2006 08:14:04 -0700

Australian Prime Minister John Howard Thursday also rejected the report. "It's not plausible, it's not based on anything other than a house to house survey."

That's pretty funny. What else would count, other than a house-to-house survey?

So Saddam's reign of terror killed 400,000 people. Bush's... well, whatever killed 650,000 people. Lawrence will now reply the 650,000 people died so they could live in freedom.

Julie asks about statistics.

Statistics is a tool in biology, physics, and other sciences. If Bush and his friends reject statistics, well, they also reject the mathematics that underlie many sciences.

I wrote about this elsewhere:

"Statisticians have shown that 664 events are sufficient to get a representative sample of a population with a 5% margin of error. If you use fewer events, the margin of error will be larger. You can use more events, but it improves the margin of error only slightly. For example, with 2,000 events, the margin of error is 2.5%. With 10,000 events, the margin of error drops to 1%. Of course, the more events, the better. We find that 700 events with a 5% margin of error is sufficient to make decisions with a good level of confidence."

The study by John Hopkins University surveyed 1,849 households, which brings the margin of error down to a bit over 2.5%. If someone reported a death, the death was independantly confirmed by checking the death certificate and the cause of death. Thus by knowing how many people died in those households, we can extrapolate to the population of Iraq and estimate how many people have died in Iraq due to Bush's attack.

For Bush, Australian Prime Minister Howard, and others to reject the study outright is simply coverup.


To change your Lit-Ideas settings (subscribe/unsub, vacation on/off,
digest on/off), visit www.andreas.com/faq-lit-ideas.html

Other related posts: