MSNBC's Hardball again falsely suggested McCain holds statistically significant lead over Obama among white suburban women
SUMMARY: Chris Matthews again aired an on-screen graphic that falsely suggested that Sen. John McCain's lead of 44 percent to 38 percent over Sen. Barack Obama among white suburban women in a recent NBC News/Wall Street Journal poll was statistically significant. A chart that appeared on-screen provided only the margin of error for the survey as a whole -- 3.1 percentage points -- and not the margin of error of 9.34 percentage points for the subset of white suburban women.
On the June 16 Hardball, MSNBC host Chris Matthews again aired an on-screen graphic, based on an NBC News/Wall Street Journal poll conducted June 6-9, that falsely suggested that Sen. John McCain's lead of 44 percent to 38 percent over Sen. Barack Obama among white suburban women was statistically significant. Specifically, a chart that appeared on-screen provided only the margin of error for the survey as a whole -- 3.1 percentage points -- and not the margin of error of 9.34 percentage points for the subset of white suburban women. Media Matters for America previously noted that on the June 12 edition of Hardball, Matthews aired two separate on-screen graphics based on the same survey that also falsely suggested McCain holds a statistically significant lead among the sample of white suburban women, without providing the actual margin of error for that subset.
While the chart was displayed, Matthews stated: "I want to start with some poll numbers here. The latest NBC News/Wall Street Journal poll has Senator Obama beating Senator McCain among women by 19 points, among white women by seven points. But McCain leads among suburban women by six." Matthews did not note that the margin of error for the subset of white suburban women was three times the 3.1-percentage-point margin of error for the poll as a whole, and the results for white suburban women highlighted on the chart were well within the margin of error and not statistically significant.
Huffington Post reporter Seth Colter Walls wrote on June 12 that after Huffington Post requested NBC's polling numbers on suburban women, "MSNBC has now provided The Huffington Post with more information on its 'suburban women' finding showing a 44-38 McCain lead over Obama. 'This is within the margin of error of 9.34 percent based on a sample size of 110 within the larger poll,' an MSNBC source wrote over email. (That's three times the margin of error for the entire poll.) This means McCain's 44 percent figure of support among suburban women could actually be as low as 35 percent, while Obama's 38 percent figure could rise as high as 47 percent -- assuming a 95 percent confidence interval (for the stat wonks in the house). Alternatively, McCain could be leading Obama 53-29. While those distant outcomes are less likely true than NBC-WSJ's 44-38 finding, that broad variance raises questions about the statistical usefulness of this one particular crosstab, as opposed to the rest of the NBC-WSJ poll on the whole."
From the June 16 edition of MSNBC's Hardball with Chris Matthews:
MATTHEWS: I want to start with some poll numbers here. The latest NBC News/Wall Street Journal poll has Senator Obama beating Senator McCain among women by 19 points, among white women by seven points. But McCain leads among suburban women by six. You first, Heidi [Harris, radio show host]. What way is this going?
HARRIS: Well, I think it's going to be interesting to see. And, to me, the biggest issue's gonna be, on the Election Day, how many people don't show up to the polls?
Can someone tell me how to post pictures on here? Including what file format to use? I tried the other day but I couldn't get it to work. I was assuming that a .PNG file format was necessary, but apparently not.
For starters, the picture has to be hosted somewhere on the Internet - I usually use Google's image search to find the image I want. Once you find it, right-click on the image and select Copy from the pop-up list. Then go the the reply window in MMFA, right-click where you want the image to appear, and select Paste from the pop-up list.
That's how I posted these images:
Next time I'll make sure the photo is already on the intertubes.
If you want to post pictures like that, you can set up an account at Photobucket, upload them there, and then copy them from your Photobucket page to MMFA.
www.photobucket.com
Beyond the conclusion being mixed... is no one else bothered by the fact that Matthews thinks nothing of labeling the poll's findings about white suburban women as "SUBURBAN WOMEN"? Are we to assume that Matthews thinks that the term 'suburban' automatically implies 'white'?
MATTHEWS: "McCain leads among suburban women by six."
No, even disregarding the issue with the poll's margin of error, that's not what the poll indicated. According to MSNBC's original report on the results of the NBC/WSJ poll, "McCain leads Obama among white suburban women (44-38)."
Does no one else find this implicit assumption that the categories "suburban women" and "white suburban women" are identical to be unsettling?
"based on a sample size of 110 within the larger poll"
Never mind, I read slow today...
But this is why this is not trivial: not only is Matthews misrepresenting the margin of error so that it seems far lower than it actually is, but he's simultaneously making it seem as though this data was drawn from a larger demographic slice than it really was. The margin of error for using about 100 people to represent all white suburban voters was high enough to begin with: Matthews compounds the problem by conflating white suburban women with the even larger demographic category of all suburban women.
And that's the most generous interpretation of this mistake. There's the much uglier possibility that whoever created this graphic simply thought, "White, suburban... what's the difference? We'll just say suburban. It takes less space and, you know, people get what that word means."
http://mediamatters.org/static/images/item/hardball-20080612-suburban2.jpg
Not sure why they dropped it here.
In any case, for Matthews to again spend another segment of his show telling his viewers that 48 white suburban women said they'd vote for McCain is suspect at best.
Wait -- where did that screenshot come from? If MSNBC had it available and consciously deleted the term "white" then this is quite significant and should be explained. Otherwise, it appears to me that someone at MSNBC thinks that the description "white suburban" is redundant. (And perhaps at MMFA as well. Why did no one catch this when writing this post?)
How can you dismiss this as a simple snafu ("woops, dunno what happened to that word! silly me...") when a previous graphic had been made for the exact same poll that correctly identified the demographic subsets? Furthermore, it wasn't just graphics: the copy that Matthews read didn't include the correct description either.
Tommy, I really don't understand your objection. This isn't some case where people are differing over opinion: this is a case where MSNBC did not report the data from its own poll accurately. It wasn't accurate in the graphic. It wasn't accurate in what the host reported orally. It wasn't accurate about the margin of error in its polling data. It conflated two distinct demographic groups even though its previous reports didn't.
I'm curious: why aren't you the least bit interested in seeing a news organization held accountable for reporting accurately? Isn't that, you know, the job of a news medium? How am I making too much of a factual error in reporting?
Let me be clear: I'm not saying that this means MSNBC and Chris Matthews are a bunch of racists and everyone at MMFA is a racist for not noticing the mistake. I just think it's worth pointing out that this aspect of Matthews' report does not correspond to the facts as we know them. It's... what's the word... misinformation? a false claim?
At the very least a correction should be made. It may be a totally insignificant distinction in the polling data. Perhaps the numbers for suburban white women and suburban women in general just magically happen to be identical. Still, it's worth noting that these were polled as two distinct demographic groups, because that's how the polling data was produced. This is basic survey methodology for social scientists and journalists.
Are you quite finished?
I never said anything about the poll results, this was covered days ago here and the poll results should have included the much larger margin of error, I wasn't making any excuse for that.
What I was responding to was your implication that there was some racially charged motive behind "white suburban" being redundant, as you asked. As I said, I think you were reading something into it. I don't err or the side of that impression as a rule.
Tommy: Are you quite finished?
Yes, I was, and obviously so. What a strange thing to ask someone who had clearly finished his comment and was inviting response. (Somehow "are you quite finished?" just doesn't carry the same discursive punch when applied on a discussion board.)
Anyway, sorry if I ranted and rambled. In response to the rest of your comments I'll be sure to make my reply as direct and brief as possible:
I see.
That screenshot is from the June 12 broadcast and was included in the prior item:
http://mediamatters.org/items/200806130003?f=h_latest
What's really scary about this conversation is that we are all so concerned - and why. Why? Because we KNOW that America is just as full of "jump on the bandwagon" voters as it is voters who would only vote a straight party ticket. What does that tell you about the American voter and the real power of the mainstream media? As unbelievable as a 3rd Bush term might sound when it comes to the logical choice of the voters this fall - I really don't think The Dems can drop their fists for one second in this General Election. More than ever before we have to get the word out whenever another lie or exaggeration is uttered - and call the opposition on their every move. It really gets old - but the only way to fight lies is with the truth.
Just had to get that alliteration out of my system ;-)
Hey Mary, you're jonsing for some more alliterations, I can tell.
And Amy Goodman would be great!
There is another way to look at the lack of significance of this statistic. Rather than saying that it fails to meet the standard of 95% confidence that is usually used, we can ask at what level of confidence is this? It turns out that if the two candidates really have equal support among white suburban women, there is a 50% chance that the poll would have shown one with a lead of six points or more. The confidence level is thus just 50%. A flip of the coin.
Another way to look at it is to ask what is the probability that Obama really leads McCain in this area. That comes to 25% (for Obama to really lead, there is a 50% chance that the poll is off by more than six points, and within that there is a 50% chance that the poll is off in favor of McCain). Like the Lakers, I am sure Obama would like better odds; but it is not time to give up the fight.
Then there is the question, why are we concentrating so much on one statistic? Out of all the breakdowns of subgroups in all the polls, this is the only one that Matthews can highlight that McCain leads? If so, that speaks pretty strongly of Obama's broad appeal.
And, btw, among Russian American hemaphrodites with hairy backs McCain has a distinct lead of 54-46. Print it, Chris Matthews!
Another question. I wonder, especially after I read the NYTimes magazine profile on CM, whether he is vying, behind the scenes, for the "Meet the Press" host job. Wouldn't that make an interesting book?
If you break it down further to white, suburban, blonde, pain-killer addicted trophy wives, McCain's lead increases to a statistically significant 100%.
I hope this works:
Just as conservatives should not talk economics until they actualy KNOW a damn thing about it, these idiots shouldn't talk stat's until they know what the hell they're talking about.
Again, I normally don't ascribe to malice that which can be explained by stupidity, but I'm getting a little tired of these guys' incompetence helping the Republicans, hurting the Democrats and huring the country.
Give Tweety a break!!
He can't be bothered with actual journalism. Come on now!! Aren't we expecting a little much from a corporate talking head?
He's just trying to find something--anything!--to show us all how well Grampy is doing. The old guy needs a little momentum, a little push to put him over the top.
Adding " white suburban Women " shows the DESPERATION of the Cable Pravda Shows, to show McSame is leading somewhere . What's next , he's leading Obama with toothless , flag pin wearing women who watch Fox.
Their " road to the White House " show is just as rigged with S. Hayes from the right wing Weekly Standard, Blankley from the right wing Moonie paper the Times and Harwood from the WSJ , paper for the elite and oh there's Rachel, but more importantly , David Gregory controlling the message , " McCain is separating himself from Bush , on environmental issues "
Chris Matthews seems distracted and preoccupied while he's conducting his interviews. Not focused, jumping all over the place.
I know he's upset that Hillary lost. It shows.
Chris Matthews is a babbling fool. Everytime I hear someone suggest "Tweety Bird" as a replacement for Tim Russert, I cringe.
Theses guys are doing all they can to keep JohnnyMac relavant,But the debates are coming and we will see,what they know to be true and that is JohnnyMac is going to be trounced in the debates.I can't wait.