George,

Which part of "[B]Both this and the earlier study are the only ones to estimate mortality in Iraq using scientific methods" [/B](bolded in my quotation from the Washington Post above) do you not understand?

[quote=Prime95;96526]Say a scientist is doing a study of some chemical process and 4 experiments come in with a result between 30000 and 90000 and one experiment comes in at 650000.[/quote]A) It's a survey, not an experiment.

The scientific method does not consist only of experiment. From the Wikipedia article on "scientific method" at [URL]http://en.wikipedia.org/wiki/Scientific_method[/URL]:

"[B]Scientific method[/B] is a body of techniques for investigating [URL="http://en.wikipedia.org/wiki/Phenomenon"][COLOR=#0000ff]phenomena[/COLOR][/URL] and acquiring new [URL="http://en.wikipedia.org/wiki/Knowledge"][COLOR=#0000ff]knowledge[/COLOR][/URL], as well as for correcting and integrating previous knowledge. It is based on gathering [URL="http://en.wikipedia.org/wiki/Observable"][COLOR=#0000ff]observable[/COLOR][/URL], [URL="http://en.wikipedia.org/wiki/Empirical"][COLOR=#0000ff]empirical[/COLOR][/URL], [URL="http://en.wikipedia.org/wiki/Measurable"][COLOR=#0000ff]measurable[/COLOR][/URL] evidence, subject to the [URL="http://en.wikipedia.org/wiki/Reasoning"][COLOR=#0000ff]principles of reasoning[/COLOR][/URL][URL="http://en.wikipedia.org/wiki/Scientific_method#_note-0"][COLOR=#0000ff][1][/COLOR][/URL]."

Experiments are one way of acquiring knowledge. Surveys are another.

[or -- Did you mean that the other estimaters were "experimenting" (in the colloquial sense) with methods that were not as scientific as the Lancet study?]

B) There's only one result from the scientific survey: 655,000.

[quote]A good scientist immediately approaches the 650000 number with great skepticism.[/quote]Are you a good scientist?

The report includes error estimates.

The author has specifically countered the criticisms of survey methodology.

The author has a history of doing similar surveys in similar (war) situations, using the same methodology, that previously produced results judged accurate enough to be adopted as official UN estimates. (Or do you have some automatic reaction to "UN" that prevents you from interpreting that as favorable to the author?)

[quote]He will rerun the experiment several times[/quote]... except that in this case, it's a survey, not an experiment. But the survey methodology [U]has[/U] been successfully used several times before.

[quote]and look for reasons why this one experiment is such an outlier.[/quote]It's not an outlier. The report lists multiple comparisons of consistency with other measures.

If you actually read my quote from the article, did you notice the "[B]Burnham said that the estimate of Iraq's pre-invasion death rate -- 5.5 deaths per 1,000 people -- found in both of the Hopkins surveys was roughly the same estimate used by the CIA and the U.S. Census Bureau. He said he believes that attests to the accuracy of his team's results.[/B]

[B]He thinks further evidence of the survey's robustness is that the steepness of the upward trend it found in excess deaths in the last two years is roughly the same tendency found by other groups[/B] -- even though the actual numbers differ greatly" part?

[quote]He would search for an experiment using a different method that would independently verify the smaller or larger number.[/quote]It's a survey, not an experiment, and it [U]is[/U] verified by other independent measures. (Did you read the article?)

[quote]Instead, you have simply embraced the outlier as the truth - because it suits your political agenda.[/quote]Folks,

Whose behavior here seems more like his only justification is a political agenda? Prime95 or me?

[quote]So, yes, you misuse science.[/quote]You appear not to ever have learned what science is really about. Or do you just need new glasses?

[quote]What's so sad (or weird) is that you can make your same political point with any of the smaller numbers.[/quote]But all the other surveys have non-scientific aspects!

Folks,

Note that the guy claiming that I can make my same political point with any of the smaller [U](nonscientific)[/U] numbers is the guy defending the side that rejects the scientific method. Consistent, right?

Now, I'm not going to claim that I've been publicly opining for a long time that the other surveys of Iraqi mortality have not been scientific-enough. But -- have you ever seen me claim that one of the other surveys of Iraqi mortality [U]was[/U] scientific-enough for me to prominently cite it? (If so -- my memory isn't perfect! -- please point out such earlier citation to me.)

[quote]Instead you give your opposition an ultra-easy way to dispute your argument or dismiss you as an extremist[/quote]Well, it _is_ easy to dispute my argument with closed eyes, isn't it?

[quote]Here are some things I see that the two polar opposites have in common:

1) Given two conflicting pieces of evidence - the one that puts the target in the worst possible light is automatically the correct one. The other is to be ignored, downplayed, dismissed, or condemned.[/quote]So, Prime95 condemns the only scientific report.

[quote]2) Given two possible motives for a decision - always choose the one that showcases the victim in the worst possible light. Since underlying motives are not in any way provable, this is easy to get away with.[/quote]So, Prime95 ascribes my action to political prejudice, not science.

[quote]3) Given two possible ways to describe a victim's action pick the one that is most prejudicial.[/quote]So, Prime95 ascribes my action to political prejudice, not science.

[Wait! Am I a "victim" in the sense Prime95 means? I don't know. So perhaps my last two (well, my last [i]one[/i], duplicated) statements are inapplicable. Stay tuned for a possible future retraction.]

[quote]4) An almost complete blindness to the fact[/quote]See my comments above about reading the article.

[quote]Emotional hatred of the victim is so strong that it is rare that rational debate will let them see what they are doing or change their ways.[/quote]Folks,

See how many rational (and a few irrational :) attempts I've made to get Prime95 to recognize the difference between the Lancet report and the others? See how rare are signs that he comprehends? See how little he changes his drive to portray me as a Bush-hater?

[quote]You have shown over and over again all four traits. Note that because of commonality #4 I can confidently predict you won't see the resemblance at all.[/quote]Folks,

Suppose you didn't know who had posted those words. Who would you think better fitted them?

[quote]Hey, pretty cool but unintentional debating trick, #4 means any attempt to refute my observations automatically proves my point![/quote]Uh ... circular reasoning?

[quote]This may just be human nature. Should Hillary become President I predict I will engage in all 4 of the actions above.[/quote] -- showing self-knowledge. That's good.

- - -

Folks,

A) Did Prime95 present any valid, actually-scientific criticism of the Lancet report that hasn't already been refuted?

B) Are the logical fallacies getting easier, or harder, to refute?

Cheesehead, sorry, but as predicted you are blind to how your hatred influences your thought processes. I suspect further discussion will prove fruitless. I suggest we let the readers come to their own conclusions.

Folks,

I previously noted that Prime95 didn't present any _scientific_ criticism. So, does he refute my exposure of the thinness of his opposition to the Lancet report by actually presenting any criticism of the scientific merits of that report now? No.

Does he present any argument that the scientific methodology of any report showing a lower total was superior to that of the Lancet report? No.

Does he offer any plausible explanation of how the Lancet report could have been in error? No.

Does he explain how totals compiled by reporters living only in the "Green Zone" (because it's too dangerous to venture elsewhere in Iraq) were superior to totals compiled by people who actually traveled to all parts of Iraq, even the dangerous areas, to conduct face-to-face interviews? No.

Does he show that he knows the difference between a survey and an experiment? No.

Instead, he offers:

[quote=Prime95]Cheesehead, sorry, but as predicted you are blind to how your hatred influences your thought processes.[/quote]

Does he offer any evidence or example of how my supposed hatred influences my thought processes, other than that I supposedly prefer the Lancet report only because it has the highest deaths estimate (even though I never previously, before that Lancet report came out, cited the report with the then-highest deaths estimate, and I never mentioned the Lancet report when it was originally released, but mentioned it in this thread only after someone else cited the 650,000 figure and someone else expressed doubt about it)? No.

Does he present any evidence that my statements about the scientific merits of the Lancet survey are wrong, irrelevant, or emotionally influenced? No.

Does he offer an excuse that he can apply to any conclusion of mine that differs from his, without reference to any evidence? Alas, yes. All his excuse requires is that he have worked himself into a belief that I am as extreme a Bush-hater as he describes the Clinton-haters to be (and that I continue to disagree with him).

[quote]I suspect further discussion will prove fruitless.[/quote]George,

That's unfortunate. Are you setting up an excuse to ignore my future arguments and persuasions?

I'm surprised by your obstinance here, because my image of you from other discussions is that you're reasonable and you're willing to consider evidence.

I concede that my critique of, and holding up a mirror to, your "emotional" argument may have made it more difficult for you to publicly admit its flaws than it would have been otherwise, so:

Folks,

I ask readers to recognize that it will show great merit and strength if George admits the flaws of his argument later in this thread.

I'd really prefer that he show an understanding of what is meant by the scientific method, so I'm willing to keep trying to find the words to get through to him.

I am _not_ willing to speculate that further discussion will prove fruitless.

Can any of you, readers, offer suggestions?

[quote]I suggest we let the readers come to their own conclusions.[/quote]Well, fine, but that was already true. I've been offering my conclusions, but I can't make anyone else adopt them -- I can only try to persuade by explaining my reasons for those conclusions.

Cheesehead, cheesehead, cheesehead, you just can't let it go. That's not unexpected.

First off, during this thread my goal has not been to defend any Bush policies and decisions. My sole goal has been to point out the grossly over-the-top nature of your anti-Bush attacks.

Let's take just one example from above - your precious Lancet survey. You have, of course, completely misunderstood my scientific method analogy.

Your goal is to shame Iraq war supporters by counting dead Iraqis. There are many different ways to count the dead and the U.N., Iraq government, IraqBodyCount.org, and Lancet have all done the job for you. They've come up with counts varying from roughly 50,000 to 650,000.

Now what would a reasonable, rational Bush critic do with this? He might pose the following question for the Bush supporters to ponder: "Given four different studies that show Iraqi casualties of between 50,000 and 650,000 do you think this war was really worth the cost? If so, how can you place so little value on a human life?"

If you had done this you would not have heard a peep from me.

Now what would an emotional, irrational Bush-hater do with the information. He would immediately conclude that 650,000 died simply because it is the biggest number. This would be done independent of the relative merits of the studies. All reasons to believe the lower numbers would be marginalized. All reasons to believe the larger number would be accentuated. Then the Bush-hater will phrase his question thusly: "Given that it is proven that 650,000 Iraqis have died in the Iraq war, how can you possibly think the Iraq war was worth the human toll?"

True to form, you've taken the bigger number and run with it. You've declared it "reality". Given my reasoning above, I can only conclude you fall clearly in the irrational, Bush-hater camp.

In case I have not been clear to all readers, let me clarify my position on the Lancet survey. As a believer in science (and yes I did "learn what science is really about" from many required classes at MIT), the Lancet study could be correct. The Lancet study could be wrong and the other three studies could be correct. The actual number might be somewhere in between. Given the rather significant discrepancies, it would be imprudent to declare one particular number or study correct. Further studies (experiments) are required, preferably by an another independent group and/or using alternative methods if possible.

[QUOTE=Prime95;96655]Now what would an emotional, irrational Bush-hater do with the information. He would immediately conclude that 650,000 died simply because it is the biggest number. This would be done independent of the relative merits of the studies. All reasons to believe the lower numbers would be marginalized. All reasons to believe the larger number would be accentuated. Then the Bush-hater will phrase his question thusly: "Given that it is proven that 650,000 Iraqis have died in the Iraq war, how can you possibly think the Iraq war was worth the human toll?"

True to form, you've taken the bigger number and run with it. You've declared it "reality". Given my reasoning above, I can only conclude you fall clearly in the irrational, Bush-hater camp.
[/QUOTE]

This seems such an unfair characterization of what Richard has said that it is hard to know where to begin! Unfortunately, the label "Bush-hater" is usually just a cheap way of dismissing what someone says without engaging that person in rational discussion. Note that even George's argument is fallacious: "If a person were a Bush-hater, they would automatically, emotionally and irrationally, accept the larger estimate uncritically. Cheesehead supports the larger estimate, therefore he must be a "Bush-hater" and therefore is emotional and irrational." If George offered some evidence that Richard's arguments were truly irrational, he might have a point, but so far, he hasn't offered any such evidence.

[quote=philmoore;96689]"If a person were a Bush-hater, they would automatically, emotionally and irrationally, accept the larger estimate uncritically. Cheesehead supports the larger estimate, therefore he must be a "Bush-hater" and therefore is emotional and irrational."[/quote]Expansionary note to readers about Phil's statement of the fallacy: The logical fallacy here is one I mentioned above: symbolically, (A=>B)=>(B=>A). In other words, "If a person's behavior B is always a consequence of having characteristic A, that implies that anyone exhibiting behavior B has characteristics A", which is false because it is also possible to exhibit behavior B [I]without[/I] having characteristic A (unless that possibility has been ruled out, which it hasn't, here).

And then there's the separate issue of whether the person has actually exhibited behavior B.

Thanks, Phil.

[quote=Prime95;96655]My sole goal has been to point out the grossly over-the-top nature of your anti-Bush attacks.

Let's take just one example from above - your precious Lancet survey. You have, of course, completely misunderstood my scientific method analogy.[/quote]Okay, explain it more carefully.

How is comparing two surveys which use different methodologies analogous to comparing two runs of an experiment that uses the same methodology for each run?

Specifically, do you claim that the Lancet study's methodology is inferior to that of other suveys? If so, please tell us just what makes it inferior.

(Shall we frame our dispute as being between MIT's training and Caltech's training? :-)

[quote]Your goal is to shame Iraq war supporters by counting dead Iraqis.[/quote]No. My initial goal in citing the Lancet survey was to simply to explain where the "650,000" figure came from.

Then that morphed into countering mistaken assertions about it.

I don't think any Iraq war supporter has to be shamed into counting dead Iraqis. In fact, I see no reason to think that any significant fraction of Iraq war supporters does not already "count dead Iraqis" when appropriate. We all count them.

But I do object to an assertion that the count that came from what seems to be the most reliable method, which has a good "track record", is necessarily wrong just because it exceeds counts obtained by inferior methods with obvious flaws (e.g., counts by only reporters who don't venture out of the Green Zone, or counts based solely on media reports).

[quote]There are many different ways to count the dead and the U.N., Iraq government, IraqBodyCount.org, and Lancet have all done the job for you.[/quote]... using different methods, not all of which are equally reliable! Why don't you ever acknowledge that, George? Why do you treat them all as though they were methodologically equal?

[quote]They've come up with counts varying from roughly 50,000 to 650,000.[/quote]... [U]using different methods, not all of which are equally reliable![/U]

[quote]Now what would a reasonable, rational Bush critic do with this?[/quote]Do you have actual expertise in impartially answering that question?

[quote]He might pose the following question for the Bush supporters to ponder: "Given four different studies that show Iraqi casualties of between 50,000 and 650,000 do you think this war was really worth the cost?[/quote]No, that's what a nonscientifically-trained person might ask, regardless of political views.

A scientifically-trained person whose training had "stuck" would [I]first[/I] ask, "How was each count obtained?"

I find it hard to believe that MIT's training in that regard was inferior to Caltech's. (Significantly, anyway :)

[quote]If you had done this you would not have heard a peep from me.[/quote]... because it doesn't occur to you to compare methodologies behind such different counts?

Does the word "methodology" mean anything to you?

[quote]Now what would an emotional, irrational Bush-hater do with the information. He would immediately conclude that 650,000 died simply because it is the biggest number. This would be done independent of the relative merits of the studies.[/quote]So, what's [U]your[/U] excuse for ignoring the relative merits?

[quote]Then the Bush-hater will phrase his question thusly: "Given that it is proven that 650,000 Iraqis have died in the Iraq war,[/quote]Oh, then I'm not a Bush-hater (by George's standard), because I've never asked a question whose premise was that the 650,000 figure had been [I]proven[/I]

See, folks?

Although (A=>B)=>(B=>A) is a fallacy,

[U](A=>B)=>(not-B=>not-A) is logically sound[/U].

If a Bush-hater necessarily asks a question based on it having been [U]proven[/U] that 650,000 have died, then anyone [I]not[/I] claiming, even in a question premise, that the 650,000 figure is proven is not a Bush-hater. QED.

I've never claimed the 650,000 figure was [U]proven[/U]. What I've claimed is that the figure comes from the only study so far that used a (good) scientific methodology.

[quote]True[/quote]Uh ... [I]false[/I], actually ...[quote]to form, you've taken the bigger number and run with it. You've declared it "reality".[/quote]... in a casual sense, before the study's scientific merits were questioned. What I'd say [I]now that I'm trying to be more precise[/I] is just that the figure comes from the only study that's used a scientific methodology.[quote]Given my reasoning above, I can only conclude you fall clearly in the irrational, Bush-hater camp.[/quote]Wow -- you take a casual use of a word ("reality"), and construct your whole argument from that? Is that what they taught you at MIT?

If you had explained earlier that your argument was based on just that one word, we might have straightened this out fairly quickly. (But we'd have learned less, so in the long run ...)

BTW, I'd still say that the scientific superiority of the Lancet report is "reality", as far as I know.

[quote]In case I have not been clear to all readers, let me clarify my position on the Lancet survey. As a believer in science (and yes I did "learn what science is really about" from many required classes at MIT), the Lancet study could be correct. The Lancet study could be wrong and the other three studies could be correct. The actual number might be somewhere in between. Given the rather significant discrepancies, it would be imprudent to declare one particular number or study correct. Further studies (experiments) are required, preferably by an another independent group and/or using alternative methods if possible.[/quote]Folks,

I'm too tired now to give a detailed analysis of the flaws in that last paragraph. Maybe in a few hours from now. What he's written there is more reasonable than most of what he wrote earlier, but it still omits key elements about which I posted earlier.

Note how he claims to have learned about science, but he has dodged every question I've posed about how the scientific merits of the Lancet study compare to those of the others. Is that what a good scientist would do?

[QUOTE=Prime95;96655]The Lancet study could be wrong and the other three studies could be correct. The actual number might be somewhere in between.[/QUOTE]

...or not. ;)

[QUOTE=cheesehead;96760]Okay, explain it more carefully. How is comparing two surveys which use different methodologies analogous to comparing two runs of an experiment that uses the same methodology for each run?[/quote]

In my analogy, I never stated the experiments used the same methodologies. My point is that rather than embracing one particular result a good scientist takes a step back and looks for reasons for the discrepancy, certainly reruns the experiment with the anomalous result, and looks for an experiment with yet another methodology to resolve the discrepancy.

[quote]Specifically, do you claim that the Lancet study's methodology is inferior to that of other suveys?[/quote]

No, never have. Nor am I willing to concede it is correct. The whole process could have been corrupted when they recruited their 8 local surveyors. Asking for workers to do a "study on excess deaths" could have resulted in some surveyors being hired with a bias toward reporting results that showed more excess deaths. I've not seen the surveyor's questions or how they conducted their interviews. Good scientific statistical methods applied to flawed data will yield flawed results. (No, I'm not claiming that happened here, but we've all experienced surveys with flawed questions that made the outcome meaningless).

I'm not overly familiar with the methodologies in the other studies. I think one surveyed local morgues and hospitals. I suppose the Iraqi government counts death certificates. Both are valid methodologies. One does need to factor in possible sources for under-counting, and this ought to be possible. IIRC, the Lancet surveyors report 90% of deaths have a corresponding death certificate. If so, then the Iraqi government ought to off by about 10%. Yes, the government has an interest in reporting lower numbers and may not be functioning well enough to collect and count all the death certificates. It would be nice if an independent group could gain access to death certificate records to come up with another data point. If hospitals and morgues are where death certificates are issued (I don't know that to be the case), then why are the hospital and morgue surveys generating significantly lower numbers? It would be nice if the Lancet surveyors gathered data on where the death certificate was issued, whether the body was taken to a morgue or local hospital. This would provide valuable data for comparing the Lancet results with the other surveys. Finally, the Lancet survey should be rerun by a completely independent group. This is an off-the-top-of-my-head list of actions that should be taken (to the extent possible in a war-torn country). I'm sure there are many other good ideas for obtaining more accurate estimates for any under-counting and/or addressing any possible problems in the ways different surveys gather their data. The more data we have, the more likely it is we can resolve the discrepancies and come to a more accurate estimate.

Until more data is available, this scientist is not going to back any one of the numbers. At this point, all I'll concede is we have a good estimate of the lower bound.

[quote]I've never claimed the 650,000 figure was [U]proven[/U]....Wow -- you take a casual use of a word ("reality"), and construct your whole argument from that?[/quote]

If I cannot use your actual words to discern what you are trying to say, then how can I have a debate with you? To me, "reality" and "proven" are pretty much synonymous in the context of a casual Soap Box forum.

[quote]BTW, I'd still say that the scientific superiority of the Lancet report is "reality", as far as I know.[/quote]

Let's just say I'm a bit more prudent in assigning the word "reality" to a particular situation.

[QUOTE=ewmayer;96774][quote=prime95]The actual number might be somewhere in between.[/quote]...or not. ;)[/QUOTE]

:smile: You got me, the number could be higher. :smile:

Just a quick note about one-and-a-half points:

[quote=Prime95;96782]In my analogy, I never stated the experiments used the same methodologies.[/quote]Ah, now that you comment on actual aspects (I almost wrote "... on reality" :), we can make progress!

(Was it the MIT-vs.-Caltech gimmick that snapped you out of the rut?)

Thanks for the correction, George. Twice I misinterpreted what you had written in your 19 Jan 07 04:31 PM posting.

(See? I'm willing to admit my mistakes, once someone points out an actual mistake instead of an imaginary one.)

[quote]My point is that rather than embracing one particular result[/quote]It only looks that way because there's only one result using that methodology. I'm favoring a methodology, not the result per se.

And ... as has been pointed out, certain aspects of the Lancet survey results are compatible with comparable portions of other surveys, so the Lancet figure is not standing entirely alone.

[quote]a good scientist takes a step back and looks for reasons for the discrepancy,[/quote]... which I did --> the differing methodologies.

[quote]certainly reruns the experiment with the anomalous result, and looks for an experiment with yet another methodology to resolve the discrepancy.[/quote]But here we run into the same "It's a survey, not an experiment" difficulty, so the analogy is still flawed.

- - -

That's all I have time for now. More later.