Demonstration of understanding of concepts: F

Possible salvation question, if answered correctly:

The same polling methodology was used to show that a figure changed by two points, with a 3% margin of error. Which of the following statements is the most likely:

(1) The real number is unchanged.

(2) The real number changed in the direction indicated by the difference.

I took calculus and stat too many years ago and I’m too rusty now to remember much more than the concept without a book by my side and I’m not near one. But a two point difference would be well within the margin of error of any poll I’m aware of.

Yes. Do the words “Gaussian distribution”, “probability density”, or “standard deviation” mean anything to you?

OK, how about “confidence interval”? In particular, if you have a poll with a 3% margin of error from about 1800 respondents, with a confidence of interval of 95%, what’s the probability that the resulting figures are *more* than 3% off? Or less than 2%? Less than 1%?

Confused? Try

http://en.wikipedia.org/wiki/Margin_of_error

And particularly the section, “The use and abuse of the margin of error”. I suspect you need it, because your retort makes you sound like you suffer from a common misperception: that if a poll reports a margin of error of +/-3%, that means that any number within the range of +/-3% of the reported figure is equally probable.

So DID Bush get a 2-point bounce from his speech and the leaks beforehand? From the AP/Ipsos poll, there’s a smallish probability that it was zero (or even less), just as there’s a smallish chance that he got a *four*-point bounce or greater. If you wanted to bet on the proposition “Bush’s support numbers on Iraq just improved”, it would be a good bet. You wouldn’t bet the farm, but so what? You should almost never bet the farm anyway.

I assume (wrongly, I guess) that whenever you cite a poll, people take statistics into account, so that when I write something like “19% of Americans strongly support Bush’s ‘surge’”, they can translate “19%” as “probably very close to 19%”. If I should be more careful, then so should you: you should not only report the margin of error, but also the confidence interval if you can find it. And always use the words “probably” and “approximately” where they are statistically appropriate.

Marc I really don’t think you and Sol Landau see eye to eye on anything or am I reading his COUNTERPUNCH articles on Fidel Castro and Hugo Chavez incorrectly.

Here’s the Dems chief function: to tell the Repubs when they’ve gone ‘too far’, as in, we can’t have a surge – let’s just keep the present troop and carnage levels.

The usual way:

http://en.wikipedia.org/wiki/Kirkuk-Ceyhan_Oil_Pipeline

Turkey is adamantly opposed to Kurdish sovereignty, de facto or de jure, and absolutely will not put up with it — until major politicos and media figures have been bribed enough to shut up about it. Would it cost a lot to bribe them enough, with million per day of Kurdish crude oil crossing Turkish territory? I think not. Is Turkey corrupt enough? Easily — it’s just a notch above Russia by typical measures:

http://www.transparency.org/news_room/in_focus/bpi_2006

Do we have to do the bribing? No — we can leave all such potential embarrassments to oil-rich Iraqi Kurds, who will also offer to continue their support of Turkey’s efforts to curb the PKK. Kurdish/Turkish antipathies are far from absolute — Turkey even had Kurdish president not so long ago. There are still some scary political minefields to be crossed, the Turkmen population of Kirkuk not least among them. But Iraq’s Kurdish leadership has crossed much more dangerous minefields in the past, and has learned some lessons.

You might indeed find some of my work archived on Z-net but anything you o find there has been pilfered. I have never written a word for them, so what u are seeing are uncompensated, lifted reprints. I think it’s more accurately called intellectual property theft.

As to fifth-rate scholar Ed Herman. I hold no personal grudge against him. I just rate his credibility as zero. He willfully misrepresented my views and those of others I know with absolutely no attempt to check with us or further verify what we said or didnt say. He deliberately slandered Saul Landua, a veteran radical, merely because Saul didnt share Herman’s lunatical views of Pacifica Radio. No matter that Landau actually had a show on said radio and Herman for his part lived outside of any Pacifica signal area. Herman’s a dottering, sloppy ideologue with no respect for facts. For one example, he has several times identified me as a supporter of the war in Kosovo when in fact I organized and broadcast one of the largest anti-Kosovo teach-ins in the country. It’s nothing personal.Just lack of respect for his shoddy work.

Actually, consulates are not necessarily sovereign territory. I believe there are exceptions, but it’s unlikely that Iran’s Irbil consulate would be one of them.

I think, however, this is sadly tied to the utter inability to attach the slightest responsability to our hugely compromised War machine. Consider what Bush has been able to do: Tell us he will listen to his commanders in the field, and when they don’t tell him what he wants to here, he fires them and brings in some who will! Yet we are told, with a telling emphisis, what stand up guys they are.

And the liberal Dems dare not say a word; for they know if they do they’ll be run through the meat grinder by the “liberal media.” This country is getting to be like prewar Japan; so cowed is the general public by the Military.

I am quite certain that our genial host-gatekeeper will be able to judge the merits of Mr. Herman’s analysis, in the spirit of anti-war unity, unencumbered by any personal or ideological differences which should surely take a backseat to the human tragedy caused by this war.

Like Mavis, you need to stop confusing lack of support from acquiesence among the Dems. Those are just tactical differences that you’re obsessing over – both party’s overriding goal remains: O_peration I_raqi L_iberation.

Of course Joe Lieberman is now the WH’s favorite Democrat (er “Independent Democrat” that is) but one reporter from ROLL CALL told MSNBC that he was “Irrelevent.” I bring this up because Mitch McConnel is threatening a filibuster which one supposes “Holy Joe” would support but there are now enough GOP Senators onboard to overide. But there are two points here:

1. Dems can prevail by refusing to fund the Supplemental that Bush is asking for to finance his adventure. Can’t filibuster that.

2. Paul Begala had a good idea – I know it happens now and then. Instead of a non-binding “sense of the Senate” Resolution condemning the war put up the President’s speech last night for an up or down vote. See who wants to escalate and who doesn’t and go from there. Put the GOP legislators – and the Dems – on record.

I sense an increased spineiness in the Senate and House to challenge this President. More are urging funds cutoffs for the surge if not for the whole show. And that was sure some reception Condi got today at Foreign Relations!

BTW Anyone seen Hillary around? She’s the frontrunner?

“Absence of dissent does not equal support”. It signifies acquiesence which is what the Dems have been doing all along.

Please see Edward Herman’s excellent analysis (Z Mag.org) of why the Dems have niether the inclination nor the ability to mount any meaningful opposition to the war (also some insightful articles by Mr. Cooper in the archives of that website!).

In regard to you Kurdish petrostate exit strategy. It would be a landlocked state. How would they get the oil out?

Yes. Do the words “Gaussian distribution”, “probability density”, or “standard deviation” mean anything to you?

OK, how about “confidence interval”? In particular, if you have a poll with a 3% margin of error from about 1800 respondents, with a confidence of interval of 95%, what’s the probability that the resulting figures are *more* than 3% off? Or less than 2%? Less than 1%?

Confused? Try

http://en.wikipedia.org/wiki/Margin_of_error

And particularly the section, “The use and abuse of the margin of error”. I suspect you need it, because your retort makes you sound like you suffer from a common misperception: that if a poll reports a margin of error of +/-3%, that means that any number within the range of +/-3% of the reported figure is equally probable.

So DID Bush get a 2-point bounce from his speech and the leaks beforehand? From the AP/Ipsos poll, there’s a smallish probability that it was zero (or even less), just as there’s a smallish chance that he got a *four*-point bounce or greater. If you wanted to bet on the proposition “Bush’s support numbers on Iraq just improved”, it would be a good bet. You wouldn’t bet the farm, but so what? You should almost never bet the farm anyway.

I assume (wrongly, I guess) that whenever you cite a poll, people take statistics into account, so that when I write something like “19% of Americans strongly support Bush’s ‘surge’”, they can translate “19%” as “probably very close to 19%”. If I should be more careful, then so should you: you should not only report the margin of error, but also the confidence interval if you can find it. And always use the words “probably” and “approximately” where they are statistically appropriate.