Is CNN biased or just mathmatically challenged?

[dsc]

http://www.cnn.com/ELECTION/2004/primaries/pages/states/NH/

RACE STATUS CANDIDATE VOTES VOTE% DEL PRECINCTS LOCAL EXIT POLL
New Hampshire
updated: 3:12 p.m.,
January 28
Kerry 84,229 39% 13 100% reporting
Dean 57,788 26% 9
Clark 27,254 13% 0
Edwards 26,416 12% 0
Lieberman 18,829 9% 0
Kucinich 3,104 1% 0
Sharpton 345 0% 0

The total voting is 217,965.

Percents
Kerry 38.64
Dean 26.51
Clark 12.50
Edwards 12.12
Lieberman 8.63
Kucinich 1.42
Sharpton 0.16

By what standard of rounding does one round Dean's total down but Clark's total up? Inquring minds want to know.


On edit.

Evidently 1,268 people voted for other. That makes the total 219,233. So now the percentages have to be adjusted accordingly.


Kerry 38.41%
Dean 26.36%
Clark 12.43%
Edwards 12.05%
Lieberman 8.59%
Kucinich 1.42%
Sharpton 0.16%

Admittedly the problem is now with Kerry and Clark. No one should have been rounded up except Lieberman. The rest of my post is accurate.

[Cuban_Liberal]

1. Good catch!

:)

[quaker bill]

2. Both.

I am sure of it.

[arcos]

3. very interesting!

This is a MUST read.

[papau]

4. Not biased - just solution to problem of 3 can be rounded up and get 100%

but 4 entries qualify for rounding up.

Usual solution is to round correctly, then say that it adds to more than 100% because of rounding.

hmmmm.... maybe it is bias.....

[dsc]

5. for your explanation to make sense

then Clark shouldn't have been rounded up.

[papau]

18. Must use same rule for all to avoid bias - so the end result adds to 101%

why a choice was made to round clark up - I guess it was needed to show that he won 3rd.

But then why Dean was the one of the remaining 3 to not get rounded correctly - well - bias is as good an explanation as any other~

:-)

[ProudToBeLiberal]

6. very observant

[carpetbagger]

7. You missed some votes, that's why your %'s are off.

CNN didn't list the other people who got votes.

Gephardt got 300+ votes, Bush got about 100 write-ins in the Dem primary, Braun and Larouche also got several dozen each, as did about a dozen other minor candidates in the primary.

Probably enough additional votes to pull down Deans ".51" under .5 but not Kerry's ".64".

There's plenty of media tricks out there, but this isn't one of them.

[ProudToBeLiberal]

9. how about clark?

[carpetbagger]

10. Got me. n/t.

[dsc]

11. still won't work

Clark had 12.50% which rounded to 13%. Those missed votes would have done to him what it did to Dean.

I also tried it with your adjustments I used 400 for Gep. 150 for Bush 84 each for LaRouche and Braun and 100 for the rest. Percent was 26.50%.

[carpetbagger]

14. You're right (except about 26.50%)

AP has it right:

http://www.newsday.com/news/politics/wire/sns-ap-nh-dem-pres-sum,0,7387596.story?coll=sns-ap-politics-headlines

Clark at 12, Dean at 26.

"Others" got 1,268.

If it's any consolation, CNN was still saying that it was too close to call Clark for third. Yeah, right. It may not have been a stellar third-place showing, but there it was.

[dsc]

17. Believe it or not it still doesn't work

Kerry has 38.41% under your numbers.

[carpetbagger]

26. I give up. (I've already given up with CNN).

CNN seems to be competing with The Daily Show, or acting like some type of national version of the local news (murders, weather, chit-chat).

I get my news from the internet, CBS radio news, network nightly news, and McNeil-Lehrer, in that order.

[dsc]

27. I editted my original post with your numbers

bizarely it made the problem worse not better. Instead of shafting Dean in favor of just Clark they did so in favor of both Clark and Kerry. CNN has gotten pretty bad. I remember it in the 1980's and just am appalled at it now.

[in_cog_ni_to]

8. Well that royally

SUCKS! Write them an email!!!! Don't let them get away with this shit without calling them on it! :grr:

Go Wes!

[ProudToBeLiberal]

12. kick

[windansea]

13. it's cuz CNN just loooooooooooooooves Wes

:hug:

[kerry-is-my-prez]

15. Five and over is rounded up, under 5 rounded down - you guys must've

slept through math 101....

[dumpster_baby]

16. Umm...standard rule of rounding: 0.5 and over gets rounded up

below that is rounded down....

This is like 5th grade math.....

[dsc]

19. Yes

and Dean got 26.51% which should have been rounded up. BTW I have a BA in math and a MA in Secondary education with a concentration in math.

[kerry-is-my-prez]

36. yes- they should have rounded Dean up...

n/t

[arcos]

20. kick n/t

[dsc]

21. kick

This post deserves more than the one thoughtful discussion it got.

[arcos]

22. kick n/t

[Feanorcurufinwe]

23. So what? Dean lost by double digits. Rounding errors are insignificant.

[dsc]

25. That error was 1/13 of the margin

and remember 10% was considered the win line for Kerry. Thus it was fully a 1/3 of his victory over expectations. It wasn't just a one percent error.

[Feanorcurufinwe]

29. Umm you are gonna have to explain that one, please.

Are you trying to claim that Kerry didn't win by double digits?

Official state tally: http://www.nh.gov/sos/presprim%202004/dpressum.htm

[dsc]

33. No I think I was clear on the fact he won by 12 points

that is what 1/13 of the margin means.

[Toucano]

24. Your calculation is correct.

http://customwire.ap.org/dynamic/files/elections/2004/primaries/by_state/NH_Page.html?SITE=YAHOOELN&SECTION=POLITICS

Your calculated percentages match these.

CNN is lazy and maybe biased. But I can't understand why they'd be biased in favor of Clark since they've done such a hatchet job on him. I know, I know...he gave them the fuel. He's wised up now.

[George_Bonanza]

28. I think they wanted it to add up to a hundred, and...

If you add up all those numbers rounded correctly, you get 98%, so you need to round some people up. Kerry and Clark are closer to the 0.50 mark than Dean, so my assumption is that they were chosen to be rounded up to make the addition equal to a hundred. Plus, it would give Clark a more solid acknowledgement of 3rd place, rather than having to explain Clark got 0.4% more votes than Edwards while the screen shows a statistical tie.

[arcos]

30. not true...

Kerry is well over 0.50 mark, and he was correctly rounded up. Clark is at 0.50 and was correctly rounded up. Dean is at 0.51 but was rounded down. How can you round 0.50 up and 0.51 down??

on edit: I hadn't seen dsc's edit, so everything on this post is invalid. Please disregard. :P

[Feanorcurufinwe]

31. Here is the official state tally

http://www.nh.gov/sos/presprim%202004/dpressum.htm


Someone who cares could use the actual numbers to figure out the exact percentages, without rounding.


But why?

[dsc]

34. But it should read

that Kerry, Clark, and Dean were all under .5 and while both Clark and Kerry were rounded up with Dean rounded down. That actually is a worse problem. I think they used my first set to round not my second but I can't be sure.

[Zomby Woof]

32. Kerry won

No matter how the numbers crunch, he soundly defeated Dean by a double-digit margin. No amount of parsing in tenths or hundredths of a point is going to change the degree of the trouncing.

[dsc]

35. If the stakes weren't so high

I would be laughing my ass off when this shit happens to Kerry and people like you then start bitching.