Cluster Points
posted at 10:23 am on October 18, 2006 by Bryan
Send to a Friend |
Share on Facebook | printer-friendly
OpinionJournal publishes the best deconstruction yet of the Johns Hopkins Lancet study on Iraqi mortality due to the war. It is not based on denial or wishful thinking, but on the thing most glaringly absent from the study: science.
The group–associated with the Johns Hopkins Bloomberg School of Public Health–employed cluster sampling for in-person interviews, which is the methodology that I and most researchers use in developing countries. Here, in the U.S., opinion surveys often use telephone polls, selecting individuals at random. But for a country lacking in telephone penetration, door-to-door interviews are required: Neighborhoods are selected at random, and then individuals are selected at random in “clusters” within each neighborhood for door-to-door interviews. Without cluster sampling, the expense and time associated with travel would make in-person interviewing virtually impossible.
However, the key to the validity of cluster sampling is to use enough cluster points. In their 2006 report, “Mortality after the 2003 invasion of Iraq: a cross-sectional sample survey,” the Johns Hopkins team says it used 47 cluster points for their sample of 1,849 interviews. This is astonishing: I wouldn’t survey a junior high school, no less an entire country, using only 47 cluster points.
Neither would anyone else. For its 2004 survey of Iraq, the United Nations Development Program (UNDP) used 2,200 cluster points of 10 interviews each for a total sample of 21,688. True, interviews are expensive and not everyone has the U.N.’s bank account. However, even for a similarly sized sample, that is an extraordinarily small number of cluster points. A 2005 survey conducted by ABC News, Time magazine, the BBC, NHK and Der Spiegel used 135 cluster points with a sample size of 1,711–almost three times that of the Johns Hopkins team for 93% of the sample size.
What happens when you don’t use enough cluster points in a survey? You get crazy results when compared to a known quantity, or a survey with more cluster points. There was a perfect example of this two years ago. The UNDP’s survey, in April and May 2004, estimated between 18,000 and 29,000 Iraqi civilian deaths due to the war. This survey was conducted four months prior to another, earlier study by the Johns Hopkins team, which used 33 cluster points and estimated between 69,000 and 155,000 civilian deaths–four to five times as high as the UNDP survey, which used 66 times the cluster points.
The 2004 survey by the Johns Hopkins group was itself methodologically suspect–and the one they just published even more so.
Read the whole thing. It’s worth your time.
Johns Hopkins is a very fine school, known for exceptional prowess in three fields: medicine (through what may be the world’s best hospital, Johns Hopkins Hospital in Baltimore), astronomy (it’s the home of the Hubble Space Telescope’s office on earth and employs some of the best researchers and professors in the field), and political science. But this study by the Bloomberg School of Public Health (its namesake is New York Mayor Michael Bloomberg, a JHU alum who has donated vast sums of money to the university) brings much discredit on such a great institution. What we’re seeing here is scientific mission creep, and it’s widespread. While the Hubble effort remains free of political taint (full disclosure–I worked there for 8 years), from earth science to psychology to public health, leftwingers are making science their advocate for political and public policy, and they’re tainting the science to do it, and their manipulation of science is by no means limited to JHU. The risk here isn’t just that we adopt bad policy based on bad information by misinforming the public, but that science itself is eventually discredited. That’s the fastest way I know of to stop progress in its tracks, and every thinking person regardless of politics should be appalled by it.
You must be logged in to post a comment.
Blowback
Note from Hot Air management: This section is for comments from Hot Air's community of registered readers. Please don't assume that Hot Air management agrees with or otherwise endorses any particular comment just because we let it stand. A reminder: Anyone who fails to comply with our terms of use may lose their posting privilege.
Trackbacks/Pings
Trackback URL
Comments
Well, it’s nice to be vindicated by a major daily. ;^)
On the 15th of this month I posted in the other thread this exact point. Further, I called what researchers did “CHERRY PICKING” — chosing data points to return a pre-determined result.
And I exposed the reason why. I wrote:
The Opinion Journal still provides a fig leaf of cover for the authors. While it is pretty hard to prove with absolute certainty that the authors were deliberately trying to foster a fraud, it it obvious to me, at least, that they cherry picked the data to suit their premise.
Lies, damned lies, and statistics, as I said. An example of circular reasoning.
georgej on October 18, 2006 at 11:03 AM
http://www.economist.com/displaystory.cfm?story_id=3352814
The Economist supports the Lancet study.
honora on October 18, 2006 at 11:23 AM
I’ll look into their basis for doing so, but I can tell you that the Hopkins study simply didn’t use enough clusters in their sample. Yes, that by itself is enough to dismiss the study’s conclusions. When you add to that that the JHU study didn’t even consider demographics–at all–the study becomes laughable.
Update: And now that I’ve read the unsigned Economist piece, I’m not persuaded by it. It relies on the support of another academic rather than examining the study on its own merits and comparing it to other similar studies and looking at their sample sizes in relation to the JHU study.
I’m sure we’ll go round and round on this since liberals tend to dismiss all evidence that doesn’t conform to their world view (still relying on the “no ties” lie, for instance, when there’s a mountain of evidence that Saddam was in fact tied to and working with terrorists), so I’ll just nip it right here. The JHU study’s authors have an anti-war agenda, and the study was designed to further that agenda. Even Iraq Body Count, another anti-war outfit that uses inflated casualty counts to mount a pseudoscientific case against the war, finds the JHU study too flawed to be believed.
Bryan on October 18, 2006 at 11:35 AM
On the surface yes it does, but they even seem to acknowledge potential problems wit the methodology stating that 33 clusters “is a relatively small sample,” and they don’t even address the democgraphic concerns.
Every attempt to calculate a total number of deaths falls within the same gnerall ballpark, and only the John Hopkins/Lancet study is far outside the norms.
In shooting, we note that when all the rounds hit the target in a tight cluster and one hits far outside of what constitutes the larger pattern, we call that a “flier” and do not count it.
The Lancet study, by any honest measure, is a “flier,” honora. Perhaps if you weren’t so politically wedded to it you could admit that the study is quite probably very wrong with both its methodology and its results.
Bob Owens on October 18, 2006 at 11:47 AM
Honora- The results make no sense. Imagine if we chose 33 clusters to determine crime in America, and randomly chose Southeast Washington, downtown Kansas City, downtown Richmond, etc. (places easy to collect data as opposed to rural areas) and then extrapolated from these clusters.
Also, I’m sure officials in these “clusters” in Iraq had to have known about the study to provide protection for the data collectors. How many households in these clusters were tipped off by insurgents or anti-Americans to give exaggerated claims?
Also what is the definition of household? Is it extended family? If so numerous people would have claimed the same person who died, thus one person would be claimed several times, really messing up results.
Last, where are all the dead bodies? It is funny how liberals will believe the Lancet study with no proof at all and then whine about no weapons of mass destruction being found.
januarius on October 18, 2006 at 11:55 AM
In shooting, we note that when all the rounds hit the target in a tight cluster and one hits far outside of what constitutes the larger pattern, we call that a “flier” and do not count it.
The Lancet study, by any honest measure, is a “flier,” honora. Perhaps if you weren’t so politically wedded to it you could admit that the study is quite probably very wrong with both its methodology and its results.
Bob Owens on October 18, 2006 at 11:47 AM
1) Your shooting example shows a real lack of understanding of stat methods and the underlying assumptions.
2) I’m politically wedded? And you are not I assume? (I supported this war therefore there can’t be this many people dead…)
3) Yeah, that Economist, what a biased rag that is. Unlike the notoriously fair WSJ. Don’t make me laugh.
4) That said, I don’t know if the Lancet article is on target or not. Neither do you.
honora on October 18, 2006 at 12:01 PM
The Lancet article with appendices is available on line. It addresses most of your questions. And honestly, do you really think any reasonably competent researcher isn’t aware of the concerns you mention? Get serious.How do you define HH? How does the Census Bureau define it–related people who live together.
I can only assume the dead bodies have been buried. Where else would they be? Is this a trick question?
honora on October 18, 2006 at 12:04 PM
I’m more aware than you seem to be that scientists can and increasingly do foist unsupportable conclusions on an unsuspecting public that often doesn’t read past the headlines. And having spent 8 years at JHU I’m more than passingly familiar with campus politics.
The JHU study is, what, about 8 pages? Based on 33 clusters to represent 27 million people, taking into account exactly no demographic information? And it’s off from all the other studies by a factor of about 1200%.
Get real. It’s not valid.
Bryan on October 18, 2006 at 12:10 PM
Bryan, I have daily contact with a guy who has a PhD in statistics and is a professor of Statistics. Would it be useful or helpful for me to ask him some questions as to what he thinks about the methedology (sic) used here?
EFG on October 18, 2006 at 12:18 PM
What are the other studies you mention?
And I spent quite a few years in market research. It was not at all unusual to test market a new product in a handful of markets, 4 to 8 usually. There may be other issues with this methodology, but claiming that the number of clusters ipso facto renders the projection invalid is not one of them IMO.
honora on October 18, 2006 at 12:20 PM
I’m not suprised, and have attributed that ’science’ to being akin to throwing sticks into the air, then interpreting the results.
The political correct machine interprets all facts that don’t support their liberal ideology as being Republican rhetoric. That explains why there were so many PEST sufferers after the ‘04′ elections, and why there are so many Americans suffering from BDS today.
DannoJyd on October 18, 2006 at 12:32 PM
Is charley rangle part of the Johns Hopkins Lancet study (shark story)? Is that where he got his information? Just curious.
tormod on October 18, 2006 at 12:36 PM
The Economist piece started losing me in the first para:
Iraq Body Count, which does just that, currently has their range from 43937 to 48783.
So I looked at the date, and saw that it was 11/4/04. Then I chuckled a little and closed that tab.
Pablo on October 18, 2006 at 12:36 PM
Your opinion is pretty consistently wrong honora. As for the other studies, they’re mentioned in the article I linked at the top of this post. In fact, several are in the blockquote in the post itself. Read it? If not, it might be a good idea to give it a little look-see. They’re named right up there if you dare to look.
EFG–yeah, sounds like a good idea.
Bryan on October 18, 2006 at 12:38 PM
Product test marketing methods could not possibly be more irrelevant to counting corpses. There’s really only one flavor of dead, and we’re all buyers at some point.
Pablo on October 18, 2006 at 12:39 PM
Yes I have read it. And I politely asked you for some clarification. Politeness being pearls to swine in this case. Ok, listen up kid: you reference a 1200% swing from the Lancet data versus “other studies”. Well, the central value of the Lancet projection is 98,000. The UNDP is 24,000. Now I may not have your deep understanding of statistice, but I know that 1200% is not the same as 300%. My conclusion was there are other studies from whence you pulled that number.
honora on October 18, 2006 at 12:48 PM
A “market” can be a huge number of people, and any competent market research would take into account things like demographics. The lancet study is an advocacy piece, plain and simple, not an attempt to get at the truth. The goal of the study was to provide such a massive number that could be cited endlessly in media articles, and even if the study was eviscerated, its defenders could still claim the ‘real’ body count was nonetheless in the hundreds of thousands. You know what? The propaganda worked. Allahpundit succumbed to it, and adopted the nonsensical view that “the truth must be in the middle” which is exactly what the folks behind this study were counting on when they grossly inflated their numbers.
Bryan is right: Iraq Body Count is an existing untrustworthy, strongly biased source that inflates the number of dead in order to push a political agenda. IBC’s numbers should be viewed as a CEILING not a FLOOR.
The projection is absolutely invalid. In SCIENCE, if there is any problem with the methodology, the conclusion is invalid. In SCIENCE, the burden of proof is on the people conducting the study to show their work. The only REASONABLE conclusion to take away from this study is that it is flawed and it provides nothing whatsoever of value. If you believe in this study despite the fact that:
(1) a previous lancet study similarly ridiculously inflating the body count was debunked several years ago, and
(2) iraq body count provides already-inflated body count figures,
then I think you are not a reasonable person, and I dont give any weight to the opinions of people who arent reasonable. Your track record on these boards has been that of a liberal partisan anyway, and this is just more of the same from you.
kaltes on October 18, 2006 at 12:49 PM
The Economist piece started losing me in the first para:
Not everybody agrees. Adding up the number of civilians reported killed in confirmed press accounts yields a figure of around 15,000.
Iraq Body Count, which does just that, currently has their range from 43937 to 48783.
The original Lancet study was done in 2004 as was the UNDP study. The latest Lancet study utilized the same methodolgy. Thus the relevance.
honora on October 18, 2006 at 12:50 PM
Statistics are statistics and sampling is sampling.
honora on October 18, 2006 at 12:51 PM
Honora, what you just mentioned is a technique for marketing. But you have to be careful about how you use cluster sampling. Cluster sampling can be effective, if used properly.
For example, suppose I wanted to see what the average reading difficulty level of a book was. I don’t want to read the whole thing. So I could select one page at random and use the words on that page to asses the difficulty. And if the book has a consistent reading difficulty level, that should be good enough. If the book gets harder as you continue reading, randomly selecting one page from randomly selected chapters should take care of that. So in this case, a book that may have a population that consists of a thousand separate pages, you could use one cluster and get an accurate result.
But clusters can bite you in the ass if you use them wrong. Let’s say we have a yellow pages phone book. And we want to find out how the economy is doing. We could call local businesses and ask them if they plan on hiring new workers. If a lot of businesses say yes, it would indicate the economy is doing good.
So what could happen if we randomly selected a page in the Yellow pages and called every business on it? Would we get a reliable result like we did with our book reading difficulty level? Probably not. This cluster will probably only contain listings for one or two business types, for example, used car dealers. And just because used car dealers are hiring, does not necessarily mean that software developers, hairdressers, and local retail stores are doing well. And since used car sales usually tend to go up when the rest of the economy is doing poorly, this could give us a completely erroneous view of the economy. So in this case our cluster was poorly selected and led us astray.
EFG on October 18, 2006 at 12:51 PM
honora said:
No, you were not polite, you stated “What are the other studies you mention?” That is neither polite nor particularly rude. In response, Bryan pointed out that, had you actually read his post, you would already know the answer to your question. His point is valid, you obviously did not read his post. The other studies figure prominently in it.
Whereas your previous comment was neutral, not polite, this comment is unmistakably rude.
Yes another rude comment from you, so if Bryan is a ‘kid’ does that make you an old lady?
The central value of the RECENT Lancet study, the study that is the subject of this whole discussion, is 98,000? Wrong!
Maybe you should get your own assumptions sorted out before you go attacking people. The lancet study at issue HERE does not project 98,000. The article on which Bryan’s entire post is based is titled:
655,000 War Dead?
THAT is the projection Bryan was referencing with the 1200%, and anyone with so much as a whit of common sense would see that very clearly had they read his post before trying to argue with him about it.
kaltes on October 18, 2006 at 1:00 PM
EFG–thank you for the civilized answer. I understand what you are saying–the issue is, is the cluster technique utilized in the L study really analogous to your example? One page from the 800 or so in your typical yellow pages versus 47 cluster points from a universe of what? What are the demographic metric units that are germane?
What is of more interest to me re the population of Iraq is the latest report that populations are shifting–Sunnis to Sunni strongholds, Shiites to Shiite strongholds. While a pretty grim indication of what’s happening there, maybe the germ of how this can eventually be settled. Perhaps.
honora on October 18, 2006 at 1:03 PM
OK Bryan, I’m printing out the Lancet study now.
If you have any question in particular you want addressed, let me know.
Otherwise, I’ll just ask him what his opinion of the methedology of this study is.
EFG on October 18, 2006 at 1:06 PM
Test marketing is glorified opinion polling. A body count is a collection of data.
Pablo on October 18, 2006 at 1:07 PM
Thanks EFG. I think just a general opinion of the study centering on the sample size, and maybe contrast it with the others that are in fact named in the article I linked and quoted would be useful. Also, some background (nothing personal) on the person you’re talking to would be useful, just so we can assess credibility. You can email any other questions and the interview’s results to me at jybmail–at–gmail–dot–com. Die, spambots!
As for honora, you’re a pretty solid example of why I’ve just stopped trying to reason with liberals. I point out information that is available to you without my having to hand-hold you to get it, and you respond with name-calling. It’s completely fruitless to even argue basic facts when such tactics are de rigeur.
Bryan on October 18, 2006 at 1:11 PM
No, it didn’t. They admit methodological errors in the last study that they claim to have addressed. Furthermore, you said that the Economist supports this study, and we’re supposed to know this because of an article printed two years before the study was done?
Pablo on October 18, 2006 at 1:13 PM
No actually test marketing involves actual product in actual retail being purchased. “Glorified opinion polling” I assume you are using to describe preference or perhaps concept testing.
Both of which involve collecting data and projecting it. The latter has more caveats as opinions don’t necessarily translate into actions. In a market test, you are measuring actual behavior. The point being that the underlying assumptions re methods of data collection to allow for valid projection are the same for market research and the type of population studies discussed here.
honora on October 18, 2006 at 1:15 PM
The older lancet study was thoroughly debunked back in 2004. The author of the study admitted to releasing the study to influence the US Presidential elections.
kaltes on October 18, 2006 at 1:16 PM
Let me help you: the 1200% number is from the WSJ article wherein the author states
A margin of error of 1200%. This tells me all I need to know about the author’s familiarity with statistics.
And just another thought, perhaps you would have better results reasonings with liberals if you eliminated statements like
and
And then of course, I am the one who is “name calling”. SSDD.
honora on October 18, 2006 at 1:29 PM
The decision to purchase a product is a personal one, ie; a matter of opinion, a preference of one over another or over none at all.
Death is…well, mostly involuntary.
Pablo on October 18, 2006 at 1:45 PM
Absolutely. What does that have to do with the statistical methods? Math doesn’t recognize content, numbers are numbers. Either you prefer Coke or Pepsi. Either you’re dead or you’re not. Doesn’t matter what incidence you are tracking, the math remains the same.
honora on October 18, 2006 at 1:49 PM
If Honora or anyone else is interested, my friend Scott at Magic Statistics reviewed the Lancet/JHU study and found it seriously wanting. Scott’s a statistician by trade, so I think his opinion carries some weight.
John on October 18, 2006 at 1:50 PM
Thanks John. I will download and look at later. Gotta run right now.
honora on October 18, 2006 at 1:52 PM
We’re all ‘involuntary buyers’ at some point. Pablo, your simple wit is very disarming. No room left to shop around :)
honora, you’re obviously an intelligent older lady, who’s quick to snap back and dive into the ‘funnel-cloud’ and swirl around wildly. In the process, what’s more amusing than all other (arguments) is your ‘demand’ for polite debate, mostly you being the knife which cuts deepest.
Not meant to lecture, just a note on an intellectual level from someone who admires your background and desire to be here, versus the 100% leftie sites, which for you wouldn’t be challenging. Name-calling and arrogance have never won anyone over. There are times for sarcasm (and those who give it must be willing to take it) but this seems too serious a topic to not grant a medium for open and polite debate. Goes both or all ways. Sincerely,
Entelechy on October 18, 2006 at 2:50 PM
When Bryan said you were “pretty consistently wrong” honora, he wasn’t name-calling, he was being charitable. You always show up with the left-wing talking points, and you always get demolished.
You used the fact that Bryan pointed out your track record at hot air as a pretext to vent your anger and level direct, personal insults at him. Is it name-calling to say someone is “wrong”? Ridiculous! But I will tell you what IS name-calling: you blatantly calling Bryan swine, and following it with “listen up, kid“.
kaltes on October 18, 2006 at 2:51 PM
Of course honora will say “Oh, but I didn’t call Bryan swine, I was saying ‘pearls to swine’”. Yes, honora, you were saying ‘pearls to swine’, a Biblical reference. Christ said to not cast “pearls before swine”, and the deeper meaning is that the pigs would not understand the value of pearls, so do not give valuable things to those who do not appreciate them
As you used it, the “pearl” was your supposed courtesy when you stated “What are the other studies you mention?” I do not believe your statement was courteous at all, and was at BEST neutral. In fact, I think it was argumentative, because your statement was pregnant with your argument that the studies that Bryan DID cite were inapplicable because those studies are not as recent as this lancet study. So, for you to say you were courteous merely because you adopted the tactic of not making the crux of your argument explicit, is nonsense.
And, to state the obvious, the “swine” in your statement was Bryan.
Because you hid your argument, and did not explain that, while you were aware of the studies Bryan cited, but that you rejected the applicability of these studies because of when they were undertaken, Bryan rightly countered that you must not have read his article, as he cited several studies and your answer gave the impression you were ignorant of them. The truth is, your attempt to bait a trap provoked the appropriate answer, and when you realized that holding back your argument made you look stupid, you lashed out with personal attacks, angry that your little ploy backfired on you, then you made your argument: the argument you should have made at the beginning instead of trying to play games.
So, in the end, honora, you end up looking not only stupid, which is what made you angry in the first place, you also appear petty and underhanded.
I suggest you argue in a more straightforward fashion in the future, and dispense with the underhanded tactics. Either that, or you can just keep on digging that hole.
kaltes on October 18, 2006 at 3:10 PM
Honora wrote:
“The latest Lancet study utilized the same methodolgy.”
“Math doesn’t recognize content, numbers are numbers.”
I have a degree in math. And I’ve have had 3 semesters of 400 level courses in statistical analysis. It is not the number crunching that is bogus here. It is the CHOICE OF DATA and the fact that one of the authors, the one acknowledged by the others to have “instigated (their word) the study, has an admitted EXPLICITLY STATED POINT OF VIEW, apriori, that makes this so-called study an exercise in GIGO. GIGO stands for “Garbage in - garbage out.”
The data selected was CHERRY PICKED. Just as the data in 2004 was cherry picked. Cherry picking data is the process of only including data that insures a particular end result. That is how you lie with statistics.
The author I refer to above, Les Roberts, was interviewed in April 2005 in the Socialist Worker — the Commie newspaper in the UK — about the 2004 “study.”
Roberts said: “but the basic idea was to find almost 1,000 households representing the whole of Iraq.”
And that is the problem with these two studies. That is where the cherry picking occurred. The quality of the results depend entirely upon the database, making the selection of the data to be of paramount importance. Pick the “wrong” households, and the data will have no connection to reality.
How do I know he cherry picked data?
Start with Robert’s own words: “I get very angry about the coverage of Fallujah…” and “The US press has been manipulated….” and “I’m disappointed that there has been no similar protest or demand for explanation in the US…” and “Most of the Democratic Party went along with this. That makes them at the least complacent in this fiasco.”
And when the Socialist Worker asked him point blank:
“I get the impression from things you’ve said that you were opposed to the war. What impact did the survey have on you personally?”
Roberts replied: “…I’m convinced that the war has been a dismal failure. People in my country might not know that for years to come. But we’ve sown the seeds of hatred to an enormous extent.”
QED.
The case for anti-war ADVOCACY has been proven.
The choice of neighborhoods, and even the number of neighborhoods and families to be interviewed, is suspect because the study authors came with a POINT OF VIEW — an anti-war point of view.
For both studies we have self admitted proof of bias, which clearly led to cherry picking data to achieve a predetermined result.
This study is garbage because it has a POLITCAL POINT OF VIEW. It is attempting to influence public policy by casting a mantle of “scientific” over what is political.
And such misuse of statistics is rather common among people hoping to influence public policy through the use of them, not just the left, FWIW.
The use of medical studies (and this purports to be a medical morbidity study) by political advocacy groups to change or direct public policy is decades old. It has been done to advocate gun control, and alter the nation’s dietary policy, to name two. In fact, John Hopkins is one of the leading institutions used to change public policy by political advocates. So their participation in anti-war advocacy is not surprising.
Remember, the “ends” justifies the “means” for the left. Whether it is to end the war or to end private ownership of guns:
The GOAL of the John Hopkins authors of the two Lancet Studies IS TO END THE WAR.
The predetermined result are two studies calculated to increase opposition to the war from “horrific” civilian casualties, and timed for release to the public in order to manipulate public opinion for maximum effect by taking the “high moral ground” of feigning “concern” about innocent loss of life.
The question for Honora and the left is: WHERE WERE YOU PEOPLE WHEN SADDAM WAS RUNNING HIS RAPE ROOMS AND HIS EXECUTION CHAMBERS?
georgej on October 18, 2006 at 4:30 PM
Bryan, the info has been sent to you via e-mail.
EFG on October 18, 2006 at 6:56 PM
I don’t want to bash Allah here, but I agree with kaltes, AP totally played right in to their hands with the “truth must be in the middle” take on the situation. I’d really just like to hear his (your if you’re reading this AP) impression after reading this? Yes, I understand that it should take something like this to actually prove it, I agree, but shouldn’t you have completely withheld judgement then? We may all have called the 650k number ridiculous for non-”scientific” reasons, but I think it’s quite clear we were right… Again, all the anti-American press, especially the terrorist networks like Al-Jazeera would have been all over it if the numbers were even close to 100k. The media loves milestones because they know the stupid public digest them easily. Anyway, I’m really just looking for your opinion on this new piece AP, because I and I think many others were disappointed in your “truth is in the middle” impression you gave off.
RightWinged on October 18, 2006 at 9:57 PM
I have an MBA from Wharton. Big deal.
Same place you were.
honora on October 19, 2006 at 11:26 AM
test
honora on October 19, 2006 at 11:31 AM
I think the hordes of people coming to sign up(like myself) have cause a bit of a traffic jam.
jematlock on October 19, 2006 at 11:33 AM