You can enlist at 17.
Yup. Except that nobody in several thousand years of western philosophy, or eastern for that matter, has been able to determine what is objectively “Right.” In the words of Cecil, I, too, think the Iraq fracas is a “damn fool war,” but banalities and platitudes like yours are just foolish, self-serving and singularly unhelpful. Wrapping your opinion in trite moral terms heightens only your self-righteous indignance. You might feel better after spouting such simplistic moral lessons, but it adds nothing of substance to the debate.
Absolutely untrue. I’ve looked at The Contract With America many times, both its small form and its book-length form. It says, paraphrased:
-They’re going to build a government that is based on Christian values.
-They’re going to spend less money and cut taxes.
-They’re going to end corruption (you’re right that this is one of the things they say)
-They’re going to get tough on crime.
-They’re going to slash welfare and base it on moral values, e.g., not giving it to underage mothers.
-They’re going to take a number of steps to legislate marriage.
-They’re going to make it harder for an individual to sue a company.
-They’re going to cut taxes on social security, thereby securing the old people’s vote.
-They’re going to slash taxes on corporations.
-They’re going to enact term limits.
Those are a bunch of specific proposals (admittedly filtered through my pinko-tinted glasses; read the link if you want the Republican filter). Only a few of them imply criticism of the Democrats; many of them are specific, visionary proposals of the legislation the Republicans promised to offer.
Compare that to www.democrats.org. What specific, visionary proposals are on tap there? Which document, do you think, has had a greater influence on elections?
Daniel
Well, let’s see what alternative solutions the Republicans have come up with so far: “The Democrats are cowards and traitors.” Yeah, that’s real constructive.
There’s a reason no one has come up with a great solution: The Bush admin has fucked us all, to the point where we’re stuck in a hopeless quagmire. To withdraw now is to admit that there were no WMDs, oil is more expensive than ever, and the Iraqis are more oppressed than ever. The threat of terrorism is at an all time high, as is anti-American sentiment. We have won no hearts and no minds. But we can’t stay in, either, because there is no hope of any sort of victory. “Mission Accomplished” was 30 bloody months ago, and we’re no closer to any sort of finish than we were then.
You complain that I have no solution. You are right, I don’t. But I’m not the one who thought up this deadly and stupid nightmare. I opposed it right from the start. Since you’re one of the morons who still supports this bloody mess, let’s look to you for the ideas. And no, “The Democrats are cowards and traitors” is not a solution.
Well, I think having a do-over of the Crusades is a damn good start.
Hey, you won’t get any argument from me that the Contract was freakin awful. I’m just saying it was effective.
Daniel
Let’s drop down into the body of that post. There we find this:
Sorry, amigo. That’s the problem with believing in polls. They sample 1000 people and try to tell you that’s what the whole country is thinking, and that’s bogus.
Now, if they had a sample size greater than a million, I might start listening. Yet you expect us to believe a poll about a country with over 281,421,906 people (2000 Census) based on a sample size of 1000.
1000 people is less than .05% of the population of Houston, TX. I wouldn’t believe a poll about Houston alone based on that sample size.
It is to laugh.
Keep telling yourself that, Hump. Whatever gives you comfort, you’re welcome to.
Then, when you’re through wallowing in denial, go learn a little about statistics theory, and come join us in the reality-based community.
Clothahump, Just a few comments and a request for clarification here.
Are you stating that you do not believe that the methodology used here is statistically valid and (assuming yes to that question) are you making the case that you in fact believe that the majority of Americans are in favor of the Iraq war and of how it is being handled?
The thing is that as far as I know, that methodology is valid and lends itself to reasonably accurate output. While I am not a statistician, I do work for a company that does surveys and the statistical analysis (via reports). This being the case, a lot of raw data passes through my hands every day.
What I can tell you concerning that is that the vast majority of the institutions that I work with are using samples that size (plus or minus a few hundred people either way) to make their business decisions. Further, in looking at the output of 1000 vs. 10,000 what I can tell you is that the basic results don’t really tend to change that much.
Again, this is based on personal experience. I have no real idea what the official “scientific” methodology is. What I will say is that it is a pretty common phenomenon (we say it with the Democrats during the last election) for people who are seeing data that they don’t like to either make claims that the sample is too small, or some other “these surveys don’t matter” excuse.
When “us” means Clothahump, you’re incorrect that I expect you to believe that the poll is statistically valid. I don’t expect you to have enough math skills to count your fingers without getting lost halfway through.
If by “us” you mean “people who have even a nodding acquaintance with basic statistical/probability theory,” then yeah, I expect them to understand the poll’s validity.
Daniel
While I know that Clothahump is wrong, I don’t really know why. Can someone explain why a sample size of 1000 is statistically valid? It seems pretty counterintuitive.
I don’ thave the math in front of me to explain it well–it’s been years since I looked at it. However, maybe this thought experiment will help.
Let’s suppose that you’ve got a giant tub of marbles. Some are solid, some are swirls. It’s about two-thirds solid, and there are, let’s say, a million marbles.
Let’s say you draw one marble out. Would it surprise you to get a swirl? Probably not: your chances are pretty good that you’ll get a swirl, even though most of the marbles are solid. They’ve been mixed together pretty well.
If you draw three out, similarly, it shouldn’t surpirse you if you get all swirls, or two swirls and one solid, or thre solids. The chances of getting exactly what you’d expect – two solids and one swirl-- are actually less than your chances of getting something dramatically different.
But if you drew out a thousand marbles, would it surprise you to get two-thirds swirls? It should: the more you draw out, the more times that probability comes into play, and the closer your estimate will get to the overall composition of the tub of marbles. Even at a hundred marbles you’re likely to be pretty close, but at 100 marbles, there’s still a decent chance that you’ll hit an anomaly.
At 1000 marbles, the odds are more than 95% that you’ve gotten to within a few percentage points of the correct estimate.
I don’t know if that helps you, but it helps me.
Daniel
That is really a whole area of science. Some Google search yielded pages that have math in them that I must admit is beyond me (One such example). What I can say is that over the years that I have spent and in looking at more raw data than I would care to think about, the instance of basic results changing in a sample size even when the number of respondents is dramatically increased is rare enough that it prompts looking for corrupted data, so-called “stuffing the ballot box” or other such problems
No, actually it wasn’t. It laid out specific policy and legislative goals, along with steps to achieve those goals. Consider the first section of the Contract:
Then it goes on to list the above mentioned bills, what they would do and why they were necessary. It was a brilliant piece of electioneering that provided a clear alternative to the way things were going and a clear vision of what the Republicans planned to do if put into office. The Democrats have absolutely nothing similar and they need to put one together before the '06 election if they hope to have any shot at making a significant dent in Republican power.
Except, of course, they have done no such thing. The Democrats may or may not mess with the tax cuts. The Democrats may or may not support gay marriage. The Democrats may or may not require the troops to come home or set a timetable for withdrawal. The Democrats may or may not support a flag burning amendment. And on and on.
To add to that explanation, think about it this way:
If you draw one marble, would it surprise you if your conclusions about the makeup of the tub were off by 67%? (I.e., you draw one swirly marble, observe that 100% of your sample is swirl, and conclude that 100% of the marbles are swirls) It shouldn’t: fully a third of the time, your sample size will be off by 67%.
If you draw two marbles, your possible combinations are:
SoSo: 2/3 * 2/3 = 4/9.
SoSw: 2/31/3=2/9
SwSo: 1/32/3=2/9
SwSw: 1/3*1/3=1/9
So now, your chances of being off by 67% are only 1 in 9 instead of 1 in 3.
Draw another marble:
SoSoSo: 2/32/32/3=8/27
SoSoSw: 2/32/31/3=4/27
SoSwSo: 2/31/32/3=4/27
SwSoSo: 1/32/31/3=4/27
SoSwSw: 2/31/31/3=2/27
SwSoSw: 1/32/31/3=2/27
SwSwSo: 1/31/32/3=2/27
SwSwSw: 1/31/31/3=1/27
So after drawing three marbles, your chances of being off by 67% are 1 in 27. After 4 drawn, it’ll be 1 in 81; after 5, it’ll be 1 in 243; and so forth.
Your chances of being off by 67% become vanishingly small very quickly, the more marbles you’ve drawn. If you play around it it, you’ll see that your chances of being off by 33% drop very quickly as well. The truth is, the more marbles you draw, the lower your chances of being off by any given percent will be. It doesn’t depend on the size of your overall group from which you draw samples (except to the tiny degree that you aren’t replacing a marble after you draw it, i.e., you’re not calling someone twice for the same poll, which makes an eensy difference to the probabilities of each successive call): it depends on the size of the sample you take.
Daniel
If you throw a coin 1000 times, you have a pretty good idea of whether the coin is fair, or weighted on one side. It’s the exact same situation. Does that still seem counter-intuitive?
Part of the reason 1000 samples is statistically valid is that survey companies go to great lengths to insure that those 1000 samples are unbiased (distributed equally among social classes, genders, age groups, etc.).
Apparently, some people still think that the size of the population directly scales against the size of the necessary sample size. Not so. 1000 is perfectly good for a nation of 40,000 or 300,000,000. What requires larger and large samples is not single questions, but multiple layers to questions and/or wanting to generalize about more and more specific groups.
Suppose that the majority of the solid marbles are in the blue bin and the majority of swirls are in the red bin. They aren’t mixed well at all. What’s to prevent me from biasing the results by picking only from the blue bin? In other words, how do the pollsters know that the marbles are well-mixed?
As someone previously said, the pollsters draw from different regions, age ranges, genders, socioeconomic groups, and so forth, to make sure they’re not all drawing from the same place. If the pollsters draw only from the red bin, then their results are accurate only for the red bin; if they draw only from the top layer of each bin, then their results are valid only for the top layer. If they draw from all over the place in each bin, their results are valid for all over the place in each bin.
Daniel
The norm, when it comes to surveys and statistical output that you report to the world, is to craft the studies in such a way as to eliminate bias and present results that are statistically valid. I see this day in and day out, even in cases where one result is desired above another.
At the end of the day, we need to also add a gut check to this. Are we to believe that there is a vast conspiratorial cabal that is out there that has created a whole discipline (that of statistical analysis) that is fundamentally flawed and that exists for no other reason than to skew any data that they get or are we going to go with the notion that an increasing number of Americans in this once case disapprove of the war and the president’s handling of it?
I will remind my friends to the Right of me on this issue of the scorn heaped on folks with weird conspiracy theories when it surrounds an issue who’s handling and results they approve of.