Sponsors

New TheBlaz poll shows 90% of other polls to be stupid Email Print

"The poll" is a time-honored method of argument, most often employed against those too naïve to know any better than to believe in the raw numbers. "Numbers don't lie!" the naive say to themselves. "Numbers are pure, sweet truth! Therefore, polls are the heavenly truths of the gods themselves!" Of course, more reasonably minded folk know better, but polls are still used to excite the masses, inflame the passions, and push the agendas of all who desire a certain end. The "poll" can be their means to that end.

Alas for all who seek the basest of truth that lies in the recesses of all polls. One would think that the method of asking a bunch of people their opinions on several topics and then coming up with the percentile data of those opinions wouldn't result in such dissembling, but here we are. Polls and their results are clouded with mystery. How does a poll come into existence? Certainly some people must have been contacted to participate in these things. Certainly questions were asked, answers were given, and numbers were spawned. Unfortunately, the numbers, while not lies, do not tell the entire story and should be viewed with a critical eye.

A recent poll I found on cnn.com (which has become a seemingly endless source of nonsense and either right-or-left partisan bullshit) shows Bush's approval rating at an all-time low. Scope out the story and then come back to me. Please, come back: http://www.cnn.com/2005/POL

ITICS/11/14/bush.poll/index

.html

The worst part of all is that seemingly intelligent, rational people have succumbed to bouts of cursing, frantic arm waving and frothing at the mouth over this. Challengers of the adminstration are so caught up in these numbers that their umbrage has risen nearly to "That-asshole-drank-my-last

-beer" levels. The most tragic aspect of all is that when they try to prove that Bush is a moron and a bad president--which should be a very simple and easy thing to do--they only cite this ONE poll, when dozens have shown that Bush has never really had what one might call "the love of the people."

Let's inspect this poll, shall we? Walk with me. I'll have you back home at a reasonable time.

(CNN) -- Beset with an unpopular war and an American public increasingly less trusting, President Bush faces the lowest approval rating of his presidency, according to a national poll released Monday.

Bush also received his all-time worst marks in three other categories in the CNN/USA Today/Gallup poll. The categories were terrorism, Bush's trustworthiness and whether the Iraq war was worthwhile.

Bush's 37 percent overall approval rating was two percentage points below his ranking in an October survey. Both polls had a sampling error of plus or minus three percentage points.

Wow! Two whole percentage points with a three-point margin of error? Yeah, that's totally huge. By the way, three girls once told me I was hot. Or maybe it was zero girls. Apparently, there's no difference.

In all seriousness, George Bush is not in "free fall mode" as many seem to think. He just sucks like he always did. The only time his approval was actually HIGH was when we just got attacked and he seemed like he would be strong about it. And he was strong. Unfortunately, he was also really stupid about it. I'm not arguing that Bush is actually approved of by most of the country. We all know he sucks. But I don't see how a 2% drop with a 3% margin of error is cause for busting out the champagne. It's really no different from last time.

Come on, people. Two points? Even without that margin of error quite possibly turning a two-point loss into a one-point gain, two points is not all that impressive when we look at the sample size this poll utilized:

"...the 1,006 adult Americans interviewed by telephone Friday through Sunday..."

Barely over 1000 people. Two percent of that is 20 people. Twenty people change their minds and suddenly Bush is in a death spiral? I guess I'm just too simple to understand. Then again, at work I routinely convince at least twenty people per day that they should order something other than what they want. It's not hard to change someone's mind, and once changed it's even easier to change it again.

By the way, I do understand statistics and I know that 1000 people is a large enough sample to accurately portray the opinions of a much larger number of people...if the poll is correctly done. Was THIS poll correctly done? What sort of sample was taken? How old were these people? Were they liberal or conservative? Democrat or Republican? How did they vote last time? What was their social class? Religion? Race? Sex? What's their geographical area? More importantly, how did people in various groups and stages of finance and geography respond? MOST importantly, did everyone get asked the same questions, same wording, same options?

In the new poll, 60 percent said it was not worth going to war in Iraq, while 38 percent said it was worthwhile. The question was asked of about half of those surveyed and had a margin of error of five percentage points.
(emphasis added by me)

So, only half of those polled were asked the above question. Which half? Did how they answer to previous questions dictate whether they were allowed to be asked this most holy of questions? These are important questions which will yield important answers. What if they only asked that question to those who mostly responded negatively to Bush? The fact that not everyone was asked that question should rasie some flags to the results of that question. Unless we know which people were asked the question and why only half were asked, there remains the possibility of skewed statistics, and in politics, often it only takes one tiny crack in the Foundation of an Argument to cause people to cease paying attention to that argument. We need people to listen, not to find one tiny POSSIBLE flaw and therefore discard the whole argument.

The fact is, we don't know who they asked those questions to, we don't know why they only asked half of the people certain questions, we DON'T KNOW.

With all of this missing knowledge, why are so many people up in arms about this poll? The level of "We're winning! This is great news!" posts I've seen is staggering in number and near-barren in their applicability to reality. That latter point is especially true for a comment I see more and more of: "If the 2004 election were to be held today, Bush would lose!"

...So? If Abe Lincoln ran for president today, he'd lose because he's dead. What does it matter if Bush would lose some fabled election today? That doesn't make him any less of a president now, nor does it remove or lessen civic responsibility to actually pay attention to what's going on in the government. It seems to simply be a feeble consolation to a group of people desperate for change, and it's about as useful.

"We're winning the country back"? Umm, not quite. You've won nothing. You gained some ground, you're making moves, you're causing a ruckus and making people think, but that finish line is still a ways away. Now is not the time to get cocky and start congratulating each other. Want to know when you've won something? When the Democrats or the Independents or the Republicans who aren't insane take control. THEN you've won (actually, then we've ALL won, though some may not be willing to concede that point). Until that point, all you're doing is talking about how cool it's going to be WHEN you win, and that sort of premature celebration is a bad idea. It's like a group of four losers sitting around and talking about how cool it's going to be at Mardi Gras and how they're going to get laid. Well, guess what? TALKING ABOUT IT DIDN'T GET US LAID!

Didn't get THEM laid, I mean...

COME ON! There are PLENTY of things to get RIGHTEOUSLY pissed about when it comes to this administration. The economy, the war in Iraq, civil liberties disappearing, the rich getting richer, scandals, lies...you don't need these polls. You don't need to be happy about random numbers and promises of future glory from your like-minded compatriots. It boggles my mind that with Bush and the government f**king up in so many different ways, polls like this are what generate the most general hatred and talking.

Prove you're smarter and better than that. Let go of meaningless and vague polls, or at least look at the WHOLE picture. Take ALL the polls and look for trends. Don't focus only on the latest poll? When you're sick, does the doctor look only at the most recent symptom?

Or does he look at the PROGRESSION of symptoms, form his opinion on all of them and THEN make his diagnosis?

This nation doesn't need people getting angry at one symptom; we need doctors.

So, please. Be doctors.


KEYWORDS: , ,

Sign up for a Complimentary Member Account... Join the community! It's fast. And it'll allow you to take advantage of all this site's great features!

< Joe Sixpack: "Why's your life so bad?... Too many illegals" | Body Politics:The Senate's Sham Rebellion Against Tyranny and Torture >
 Display:
I think you effectively and provocatively pointed to a deeply injected bias of all poll result perceptions. That is  --

  • If the poll results are in your favor, they must be correct.

  • If they're against you, they must be incorrect.

  • If there are multiple polls, the one that agrees most with you is the one that is correct.

  • Polls can be a general indication of something, but there are always more important things to focus on.

    Is that correct?

    Political Cortex -- Brain Food for the Body Politic

    by Tom Ball on 11/15/2005 02:02:58 PM EST

  • A smart conservative would not be a troll. An oxymoron, perhaps, but not a troll.

    This, however, is simply sloppy and long-winded, attacking strawmen more than actual forms of folly. And it uses relatively innocent forms of naivite to cast implicit guilt-by-association aspersions on sober consideration of polls, which can be invaluable.

    Is there a valid point, as some have said?  Yes. Is it new, fresh or original? No.  Is it even reasonably well-focused on making the point it does have? No.

    One word: Diary.

    "Be realistic. Demand the impossible!" --Wall poster from the 1968 Paris Uprising

    by Paul Rosenberg on 11/15/2005 02:34:45 PM EST

    The most tragic aspect of all is that when they try to prove that Bush is a moron and a bad president--which should be a very simple and easy thing to do--they only cite this ONE poll, when dozens have shown that Bush has never really had what one might call "the love of the people."
    I don't know any example that fits this description. Do you???

    "Be realistic. Demand the impossible!" --Wall poster from the 1968 Paris Uprising

    by Paul Rosenberg on 11/18/2005 10:04:53 PM EST

    [ Parent ]
    I don't know any example that fits this description. Do you???

    I'm assuming you're speaking of examples found on this site and other's like it? Like KOS for example? If not, I'd appreciate clarification.

    I wasn't speaking of examples found on sites like this, as I'm very new to this sort of site and posting on it. I was mainly speaking of news organizations and message boards that don't have a political theme, so mostly casual conversation. The vast majority of people who post on sites like this are extremely well-informed, research oriented, and have been for some time. However, the vast majority of the American public--which is the audience I believe sites like this strive to reach and sway--are NOT as well informed nor do they pay as much attention as the average political blogger. I feel that's a problem that must be overcome if there are to be changes that many desire.

    That's basically who that line was targetting: average people who either aren't as well informed or who don't pay attention to the news and other information sources, or both.

    Of course, everything I just typed is moot if you didn't mean what I thought you meant.

    Could you please clarify for me? I'm not seeing how what you quoted from my piece is a strawman, as by and large the casual message boards and conversations I've heard and read focous on this one poll as proof, which was my point.

    by TheBlaz on 11/19/2005 05:21:52 PM EST

    [ Parent ]
    Nor do I know how my latest Political Cortex poll -- inspired by this piece -- ended up in limbo instead of the front page.

    That said, I recommend everyone have a little fun with an unscientific poll that will help PC founders get a feel for what might be (or become) an impportant issue for the community.

    by S M Dixon on 11/15/2005 03:05:00 PM EST

    [ Parent ]
    Alas for all who seek the basest of truth that lies in the recesses of all polls.

    Do you mean that there is truth at the heart of polls but it is baser truth than you will find elswhere or that it is debased truth -- truth hat has lost its meaning?

    One would think that the method of asking a bunch of people their opinions on several topics and then coming up with the percentile data of those opinions wouldn't result in such dissembling, but here we are.

    I'm not sure who you mean is dissembling here.  Are you saying that CNN is dissembling, because they are fudging their poll numbers, that interpreters of the poll are dissembling because they are lying about its meaning, or that those rejoicing in the results are lying?  The subject here is difficult to discern.

    The worst part of all is that seemingly intelligent, rational people have succumbed to bouts of cursing, frantic arm waving and frothing at the mouth over this. Challengers of the adminstration are so caught up in these numbers that their umbrage has risen nearly to "That-asshole-drank-my-last   -beer" levels

    This passage left me confused. Are opponents of the President overly pleased by the results of this poll or are they overly angry about it?  Here you imply that they are angry, but I'm not sure what they would be angry about. That his numbers have reached their lowest point?  I would think that that would please them as it would confirm their own opinion of him was shared by a larger number of Americans than anytime previously.  It seems more logical to assume that it is his supporters who would be upset by this poll and the other 8 that have come out in the past two weeks showing approval numbers in the 30's.

    Overall, I think your point that a single poll may not be able to offer significant information in and of itself is valid, however as ColdFusion notes above, a sequence of polls that show a trend are significant.  

    I think you are right to suggest skepticism as to methodology in any given poll, but again, there are 8 polls that have come out in recent weeks indicating the same trend and obtaining virtually the same result.  That provides much stronger evidence that this poll is correct than you indicate in your analysis.

    Finally, I think your most valid point is that it is unproductive and even harmful for progressives to sit back, revel in the polls, and think that the fight is over.  The fight is just beginning.  These polls indicate that people are ready to listen to why the Bush WH has been disastrous for this country, but that does not mean that they are ready to vote for change. After nearly five years of unchallenged propaganda where the neocons and religious right have borrowed from Hitler's playbook and accused their opponents of doing exactly what they themselves are guilty of, it will be difficult to explain the truth about Democrats to moderate Republicans and Independents.  Restoring truth and eliminating the Orwellian habit of this administration in calling black white and white black will take time, effort, and determination.

    All these polls indicate is that at last people are willing to listen.

    by katerina on 11/15/2005 02:44:41 PM EST

    It's always good to remember that polls are authored. There's always a reason behind them. When it comes to polls sponsored by the news media, they're intended to...make news!

    Someone at some point figured that news outlets didn't have to go out and look for stories: they could make their own. So they take a poll and the results become the biggest story of the day.

    It's the horserace. You need numbers for a horserace. Who's up, who's down? Who tossed a football on the tarmac? Who played saxophone on Arsenio? Who looked silly riding in a tank?

    How much brush could a lame duck clear if a lame duck could clear brush?

    Anyway, whoever said, "The only poll that matters is on election day" was right.

    (But DAMN 37% sucks! haha)

    Dissent Protects Democracy

    by cscs on 11/15/2005 04:26:36 PM EST

    how to interpret a poll, I recomend a brush-up course from Mystery Pollster. I am leary about an article that cites a poll but does not provide a link to the full details from the polling orginization. When they do, most of the details you wonder about are specified.

    The other point is multiple polls indicating common trends, and the trends temselves from prior polling. Again, if provided, this is useful information.

    by roysol on 11/16/2005 08:51:35 AM EST

    There's just so much blood it can lose without becoming totally impotent.  We will not be raped much longer:  Alito going down; medicare drug plan blackhole implodes; congress moving away from stepford wife mentality. Bush can still do a lot of damage but congress has the power to take his money away.
    Bush is out of the country on a trip so that he will be too busy to hear the voices in his head.  Someone should tell him those are our voices, real voices.

    "welcome to the monkey house", 1961, kurt vonnegut

    by realheathen on 11/16/2005 03:50:29 PM EST

    New to this blog but unclear why this is a front page post. It seems rather naïve.

    Granted, there are numerous potential problems in polling, including but not limited to: sample bias (inadvertent or due to bias of polling organization), push-polling (asking questions to get a desired response)), use of ambiguous questions (e.g., what values are values questions asking about?), and interpretation of "change" or "difference" when numbers are within the margin of error. For this reason, it is important to read polls results intelligently and look at how polls are conducted. It doesn't make sense to dismiss polls with a broad stroke without addressing these issues.

    In addition, while results of an individual poll or an individual polling organization may be suspect, looking at trends over time and across polling organizations provides important information about the robustness of polling results. E.g., Bush got a huge bump following 911, a smaller bump following the Iraq invasion, an even smaller bump after Saddam was captured, and a slight bump, enough to get him elected, at election time.. But he started out with mediocre numbers and, except for the bumps has followed a downward trajectory throughout his time in office. For people not familiar with these trends and differences among polling organizations, suggest they check out Polling Report and Pollkatz .
    For basic information about polls, Pollkatz, as well as Mystery Pollster suggested by roysol above are helpful.  

    by Psyche on 11/21/2005 01:20:44 PM EST

     Display: