Comments on: Lancet study of Iraqi deaths is statistically unsound and unreliable http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/ "I accept no responsibility for statistics, which are a form of magic beyond my comprehension." -- Robertson Davies Sat, 13 Dec 2008 07:05:30 +0000 http://wordpress.org/?v=2.9.1 hourly 1 By: Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » In the spotlight again: Lancet study of Iraqi deaths http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-21916 Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » In the spotlight again: Lancet study of Iraqi deaths Sat, 05 Jan 2008 03:10:13 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-21916 [...] 2.  For example, this: The study’s description of field operations indicates that the survey was administered more quickly and more smoothly than any comparable survey I’ve ever heard of.  A team of two surveyors appeared [...] [...] 2.  For example, this: The study’s description of field operations indicates that the survey was administered more quickly and more smoothly than any comparable survey I’ve ever heard of.  A team of two surveyors appeared [...]

]]> By: Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Was 2004 Lancet study correct to toss out Falluja data? http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-18746 Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Was 2004 Lancet study correct to toss out Falluja data? Thu, 26 Jul 2007 02:56:53 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-18746 [...] Was 2004 Lancet study correct to toss out Falluja data? By StatGuy I’ve written several posts about the 2006 study of Iraqi deaths, written by Burnham et al, but not about the predecessor study, authored by Roberts et al and published in 2004.  (This blog started in August 2005.)  Both studies analysed data collected in personal interviews at Iraqi households selected using cluster sample methodology and purported to find that tens of thousands of Iraqis have died because of the US invasion.  Both were published in The Lancet. [...] [...] Was 2004 Lancet study correct to toss out Falluja data? By StatGuy I’ve written several posts about the 2006 study of Iraqi deaths, written by Burnham et al, but not about the predecessor study, authored by Roberts et al and published in 2004.  (This blog started in August 2005.)  Both studies analysed data collected in personal interviews at Iraqi households selected using cluster sample methodology and purported to find that tens of thousands of Iraqis have died because of the US invasion.  Both were published in The Lancet. [...]

]]> By: Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Lancet co-author not allowed to land in Britain http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-15527 Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Lancet co-author not allowed to land in Britain Thu, 19 Apr 2007 05:10:44 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-15527 [...] Another co-author of the Lancet study, Les Roberts of Johns Hopkins Bloomberg School of Public Health, has been asked to speak if Dr Lafta can’t make it.  So, there may still be an opportunity to ask questions about the study’s methodology. [...] [...] Another co-author of the Lancet study, Les Roberts of Johns Hopkins Bloomberg School of Public Health, has been asked to speak if Dr Lafta can’t make it.  So, there may still be an opportunity to ask questions about the study’s methodology. [...]

]]> By: Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Lancet studys field manager to visit Vancouver http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-15348 Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Lancet studys field manager to visit Vancouver Fri, 13 Apr 2007 02:22:44 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-15348 [...] The study’s description of field operations indicates that the survey was administered more quickly and more smoothly than any comparable survey I’ve ever heard of.  A team of two surveyors appeared unannounced at the front doors of 1849 Iraqi households asking highly sensitive and intrusive questions about everyone who had lived there since January 2002. Interviews were successfully completed in an average of only fifteen minutes each.  We are told that interviewers found family heads at home in all but 16 households.  What’s most mind-boggling (to me, anyway) is that only 15 households refused to participate—a refusal rate of only 0.8%. [...] [...] The study’s description of field operations indicates that the survey was administered more quickly and more smoothly than any comparable survey I’ve ever heard of.  A team of two surveyors appeared unannounced at the front doors of 1849 Iraqi households asking highly sensitive and intrusive questions about everyone who had lived there since January 2002. Interviews were successfully completed in an average of only fifteen minutes each.  We are told that interviewers found family heads at home in all but 16 households.  What’s most mind-boggling (to me, anyway) is that only 15 households refused to participate—a refusal rate of only 0.8%. [...]

]]> By: Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Freedom Of Information request needed to extract data from global warmists http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-15129 Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Freedom Of Information request needed to extract data from global warmists Fri, 06 Apr 2007 01:08:51 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-15129 [...] Now that I think of it, the authors of the Lancet study of Iraqi deaths have also refused to release their raw data.  I wonder if an FOI Act request would pry those data loose. [...] [...] Now that I think of it, the authors of the Lancet study of Iraqi deaths have also refused to release their raw data.  I wonder if an FOI Act request would pry those data loose. [...]

]]> By: Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » A specification of the Lancet study‚Äôs “Main street bias” http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-8158 Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » A specification of the Lancet study‚Äôs “Main street bias” Wed, 06 Dec 2006 04:05:03 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-8158 [...] Lancet study of Iraqi deaths is statistically unsound and unreliable [...] [...] Lancet study of Iraqi deaths is statistically unsound and unreliable [...]

]]> By: Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Lancet authors slow to provide answers http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-8006 Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Lancet authors slow to provide answers Mon, 04 Dec 2006 05:24:36 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-8006 [...] I, along with many other statisticians, expressed serious reservations about the soundness of the survey and the reliability of the results.  That The Lancet authors appear unwilling to answer basic questions that the original article’s inadequate documentation left unanswered only diminishes what credibility they have left.  The unavoidable question now: Are they trying to hide something? [...] [...] I, along with many other statisticians, expressed serious reservations about the soundness of the survey and the reliability of the results.  That The Lancet authors appear unwilling to answer basic questions that the original article’s inadequate documentation left unanswered only diminishes what credibility they have left.  The unavoidable question now: Are they trying to hide something? [...]

]]> By: CaNN :: We started it. http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-5309 CaNN :: We started it. Mon, 23 Oct 2006 16:48:31 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-5309 [...] CHRISTOPHER HITCHENS on the moral idiocy evident in Lancet‚Äôs work. Stick a fork in the Lancet study‚Äîit‚Äôs done; Lancet study of Iraqi deaths is statistically unsound and unreliable …. (timblair, magicstatistics) [...] [...] CHRISTOPHER HITCHENS on the moral idiocy evident in Lancet‚Äôs work. Stick a fork in the Lancet study‚Äîit‚Äôs done; Lancet study of Iraqi deaths is statistically unsound and unreliable …. (timblair, magicstatistics) [...]

]]> By: Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Main street bias in Lancet study http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-5288 Magic Statistics - “I accept no responsibility for statistics, which are a form of magic beyond my comprehension.” — Robertson Davies » Main street bias in Lancet study Mon, 23 Oct 2006 03:26:18 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-5288 [...] Main street bias in Lancet study By StatGuy One of the joys of blogging for me is interacting with people I'd never have met otherwise.  My posts on the Lancet study of Iraqi deaths have afforded many opportunities for that.  (Background here.)  One in particular prompts this post. [...] [...] Main street bias in Lancet study By StatGuy One of the joys of blogging for me is interacting with people I'd never have met otherwise.  My posts on the Lancet study of Iraqi deaths have afforded many opportunities for that.  (Background here.)  One in particular prompts this post. [...]

]]> By: Sceptic http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-5282 Sceptic Mon, 23 Oct 2006 00:58:04 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-5282 I don't understand how the peer reviewers can have missed all of this -- or, for that matter, all these signatories of a letter supporting the study: <a href="http://www.theage.com.au/news/opinion/the-iraq-deaths-study-was-valid-and-correct/2006/10/20/1160851135985.html?page=fullpage" rel="nofollow">http://www.theage.com.au/news/opinion/the-iraq-deaths-study-was-valid-and-correct/2006/10/20/1160851135985.html?page=fullpage</a> I don’t understand how the peer reviewers can have missed all of this — or, for that matter, all these signatories of a letter supporting the study:

http://www.theage.com.au/news/opinion/the-iraq-deaths-study-was-valid-and-correct/2006/10/20/1160851135985.html?page=fullpage

]]> By: StatGuy http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-5278 StatGuy Sun, 22 Oct 2006 21:48:20 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-5278 Sceptic, Thanks for your comments. The points you raise are all important. 1. Very good point. In both Lancet surveys, only one household was randomly chosen within each cluster. Once the starting point was selected, the rest of the households in the cluster followed in order. The ILCS surveyed ten households within each cluster, and all ten were randomly chosen 2. That is still an unresolved question. 3. Indeed, it does. I plan to have a further post on that later today. 4a. This is one of the pitfalls of charting statistical data. Charts can be useful for illustrative purposes, but they can be manipulated by altering the scales of the axes. That is another reason for relying primarily on the statistical analysis. No chart can overcome or outweigh analytical problems. 5. The Lancet study mentions collecting age data, but not birth dates. This seems to mean that, for each person who had died, the researchers simply asked the age of the person at death. They did not gather birth dates as such. Also, Moore’s article says the researchers did not gather any demographic data on living persons. <em>[W]hile the gender and the age of the deceased were recorded in the 2006 Johns Hopkins study, nobody, according to Dr. Roberts, recorded demographic information for the living survey respondents. This would be the first survey I have looked at in my 15 years of looking that did not ask demographic questions of its respondents.</em> Business surveys generally do not gather demographic data on respondents. No one wants to know the birth date of the clerk filling out a firm’s questionnaire on manufacturing shipments. But for population surveys, Mr Moore is absolutely correct. Questions on sex and birth dates appear at or very close to the top of every household or individual survey I’ve ever seen. Sceptic,

Thanks for your comments. The points you raise are all important.

1. Very good point. In both Lancet surveys, only one household was randomly chosen within each cluster. Once the starting point was selected, the rest of the households in the cluster followed in order. The ILCS surveyed ten households within each cluster, and all ten were randomly chosen

2. That is still an unresolved question.

3. Indeed, it does. I plan to have a further post on that later today.

4a. This is one of the pitfalls of charting statistical data. Charts can be useful for illustrative purposes, but they can be manipulated by altering the scales of the axes. That is another reason for relying primarily on the statistical analysis. No chart can overcome or outweigh analytical problems.

5. The Lancet study mentions collecting age data, but not birth dates. This seems to mean that, for each person who had died, the researchers simply asked the age of the person at death. They did not gather birth dates as such. Also, Moore’s article says the researchers did not gather any demographic data on living persons.

[W]hile the gender and the age of the deceased were recorded in the 2006 Johns Hopkins study, nobody, according to Dr. Roberts, recorded demographic information for the living survey respondents. This would be the first survey I have looked at in my 15 years of looking that did not ask demographic questions of its respondents.

Business surveys generally do not gather demographic data on respondents. No one wants to know the birth date of the clerk filling out a firm’s questionnaire on manufacturing shipments. But for population surveys, Mr Moore is absolutely correct. Questions on sex and birth dates appear at or very close to the top of every household or individual survey I’ve ever seen.

]]> By: Roger’s View of the World, Love and Seafood Gumbo! » Gil Burnham — “Weenie Man of the week!” http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/comment-page-1/#comment-5241 Roger’s View of the World, Love and Seafood Gumbo! » Gil Burnham — “Weenie Man of the week!” Sat, 21 Oct 2006 18:59:50 +0000 http://magicstatistics.com/2006/10/15/lancet-study-of-iraqi-deaths-is-statistically-unsound-and-unreliable/#comment-5241 [...] Major Statistics Dumb Looks Still free SEIXON Gateway Pundit Decision ‘08 America’s majority [...] [...] Major Statistics Dumb Looks Still free SEIXON Gateway Pundit Decision ‘08 America’s majority [...]

]]>