LOL, CNN Can't Count

[kpete]

[MrSlayer]

1. It says "within sampling error" so I guess that's the justification.

Blitzer is a clown.

[alcibiades_mystery]

2. Except the sampling error is 3.5%

Is 4 within 3.5?

I'll have to check my Wolf Blitzer math.

[Ms. Toad]

6. Where is the 3.5 coming from?

(there isn't a link, but my foggy brain reads 4 on the image)

[Mass]

9. The problem is that they publish results in whole number, so we have no clue what the difference

really is.

It could be 49.5-46.4 = 3.1
or it could be 50.4-45.5=4.9
Or anything in-between. Who knows whether this poll is within the MOE or not.

What we know, though, is that CNN would not have called this poll tied if Romney was leading. I think this is pretty clear.

[Ms. Toad]

11. Probably right on that one. n/t

[Ms. Toad]

13. I found a more complete report -

but still not enough detail to answer your question.

http://politicalticker.blogs.cnn.com/2012/10/26/cnn-poll-obama-50-romney-46-in-ohio/

Lousy reporting to include one decimal point in the MOE, and none in the poll results.

[dennis4868]

5. But it's not

MOE is 3.5! CNN SUCKS.

[Ms. Toad]

8. Where is 3.5 coming from? n/t

[Mass]

12. Poll results here

http://i2.cdn.turner.com/cnn/2012/images/10/26/topoh2.pdf

[Ms. Toad]

16. Thx. That's more complete than the one I found -

But still not enough to answer the question you asked above, since it does not include raw numbers, and rounds the percentage to the whole number.

[Spazito]

15. From the CNN poll itself...

"The CNN poll was conducted by ORC International from October 23-25, with 1,009 Ohio adults, including 896 registered voters and 741 likely voters, questioned by telephone. The survey's overall sampling error is plus or minus 3.5 percentage points."

http://politicalticker.blogs.cnn.com/2012/10/26/cnn-poll-obama-50-romney-46-in-ohio/?hpt=hp_t2

The other misleading part of this is assuming the sampling error if non-existant would be completely in Romney's favor.

[Ms. Toad]

17. It would certainly be more accurate to say

the range of outcomes is from a tie to a 7 point Obama lead.

[Spazito]

18. CNN has an anathema to accuracy...

since this campaign season began, imo. Blitzer has a deplorable bias in his reporting and this is but one example, unfortunately.

[Hutzpa]

3. Sad and sorry. nt

[Ms. Toad]

4. Not sure what you're getting at.

My brain is rather foggy at the moment - but 50-46=4, and the text say 4 point lead within sampling error = which makes them statistically tied (. (And I assume the other 4% are undecided or voting for 3rd party candidates.)

Am I missing something?

[MFM008]

7. 100% percent

of WOlf Blitzer is a tool.

[hrmjustin]

10. They want a close race for ratings.

[CreekDog]

14. Oh come on, Wolf can count

like how many paychecks does he get?

[Honeycombe8]

19. They keep DOING that! What is up with that? (as if I don't know)nt

[Cali_Democrat]

20. and if a poll comes out showing Romney with the lead,

people like Wolf Blitzer absolutely wet themselves as they go on and on about Romney's "momentum"

[Honeycombe8]

21. Exactly. That's the truth. I'm not imagining that. nt

[Marinedem]

22. Take a stats class

4% is the standard margin of error.

[Orangepeel]

23. The President could be at 47% or 53%. Romney could be at 43% or 49% (but he's not)

The margin of error applies to both numbers, not the difference.

BUT this isn't the only poll. People take the average of polls based on the idea that if we consider them as a whole, the sample size is much bigger and the margin of error is much smaller. If there are 10 polls and the President is ahead in all of them, he's ahead (which he is)

CNN still sucks.

[Zoeisright]

24. Leslie is a complete moran (sic).

He's carried so much water and bullshit for repukes his eyes are swimming.

[upi402]

25. CNN sucks ass. ALERT THE MEDIA!!!

wait wut?