Linear Regression GRAPH: Kerry State Vote vs. Exit Poll completion rate

[TruthIsAll]

There's a negative correlation. As Kerry's state vote percentage increases, the exit poll completion rate decreases.

Another finding which makes rBr even less plausible..



Thanks to Bill Borer for providing the following data and stats:

K Vote Compl
AllPrecincts 53.2
NationalSurvey 52.8

K Vote Compl
Utah 27.06 59.6
Wyoming 29.7 66
Idaho 30.71 63.2
Nebraska 32.53 66.5
Oklahoma 34.44 53.2
NorthDakota 36.09 63
Alaska 36.17 53.2
Kansas 36.97 64.5
Alabama 37.08 58.3
Texas 38.49 58.3
SouthDakota 39.09 42.7
Indiana 39.46 38.6
Montana 39.51 63
Mississippi 39.91 49.6
Kentucky 39.99 52.6
SouthCarolina 41.31 59.4
Georgia 41.58 63.9
Louisiana 42.63 47.8
Tennessee 42.78 66.7
WestVirginia 43.48 48.7
NorthCarolina 43.72 52.6
Arkansas 44.74 60.2
Arizona 45.03 57.3
Virginia 45.65 56.4
Missouri 46.33 47
Colorado 47.35 55.5
Florida 47.47 49
Nevada 48.67 49.1
Ohio 48.75 44.1
NewMexico 49.42 56.9
Iowa 49.54 52.6
Wisconsin 50.2 55.3
NewHampshire 50.68 44
Pennsylvania 51.13 46.8
Michigan 51.73 50.2
Minnesota 51.76 45.3
NewJersey 53.13 59.7
Washington 53.6 53.8
Delaware 53.82 57.5
Hawaii 54.37 53.4
Maine 54.48 61.3
Illinois 54.99 51.9
California 55.21 50.5
Connecticut 55.24 51
Maryland 56.25 59.4
NewYork 58.79 57.9
Vermont 60.34 53.1
RhodeIsland 60.48 44.2
Massachusetts 62.7 56.5
D.C. 90.52 53.5

KV vs. Compl
r Val. -0.278 58576
t Val. 2.01
p Val. 0.0501

[pansypoo53219]

1. numbers confuse the

gnews media. show it with a BAR graph. maybe THEN they get it.

[Zan_of_Texas]

3. TIA works hard on the numbers, but

refuses to spend any time responding to normal people's requests that the graph be explained in words, right on the graph -- what do the numbers on the left axis represent, what do the numbers on the horizontal axis represent.

Take a state and a few data points, and explain what they mean in words, as an example.

[TruthIsAll]

5. Come on, you're being a little harsh...

You:

TIA works hard on the numbers, but refuses to spend any time responding to normal people's requests that the graph be explained in words, right on the graph -- what do the numbers on the left axis represent, what do the numbers on the horizontal axis represent.

Take a state and a few data points, and explain what they mean in words, as an example.

Me:
I thought the table/graph/explanation was self-explanatory, but Ok, maybe I assumed too much. The table shows Kerry's state vote%, followed by the exit poll completion rate.

Purpose: To show the relationship between Kerry's state vote and exit poll completion rate (the percentage of those who completed the exit polls).

Graph 1: Kerry's % of the state vote, in ascending order, left to right.
Graph 2: Corresponding Exit poll completion rate, left to right.

For both graphs, I have calculated the best-fit straight line.
y = a + bx is the equation of a straight line, where b is the slope.

Both "best-fit" linear equations are clearly stated and shown as dashed lines on the graph.

Observe:
Kerry's vote percentage increases in states going from left to right (increasing slope). The corresponding state exit poll completion rates decrease (negative slope).

Conclusion:
There is a negative correlation between Kerry's vote and the completion rate. They move opposite to each other.

This contradicts the rBr hypothesis, which states that Kerry voters were more anxious to speak to the exit pollsters than Bush voters.

If rBr was a viable hypothesis, as you moved into Kerry strongholds, you would expect that there would be more enthusiasm on the part of voters to be interviewed, not less.

If rBr was a viable hypothesis, as you moved into Bush strongholds, you would expect that there would be less enthusiasm on the part of voters to be interviewed, not more.


Hope that clears it up.

[Zan_of_Texas]

6. Okay. Now we're talkin'!

[TruthIsAll]

4. It's a linear graph. Click on image. A bar graph will not tell the story..

/

[kster]

2. kick.nt

[Chi]

7. Where are the naysayers, I would think they would be all over this

Certainly appears self-evident, and puts a hurting on the rbr theory.

Does this now mean we are back to 'M/E's own info shows rbr to be highly unlikely...at best'?
Wait...wasn't there a paper that said M/E's info didn't show this...hmmm.

<Chi checks his watch and waits for a PM from the shadows>

[autorank]

8. They are licking their wounds. There is only so much of an ass kicking

any group can take before they retreat to renew their negativism.

Good point. TIA's arguments especially after "Hobson's Choice" and the major thread that preceeded it 43/37 (remember, they wouldn't to the math) are so devastating, they have nothing of substance to say.

Excellent point!

[byronius]

9. Another nail in the old pine, TIA. Keep hammering.

Sooner or later, this box goes in the ground, along with all the other gremlins who've plagued the better half through history.

[autorank]

10. KICK

:kick:

[TruthIsAll]

11. REGRESSION GRAPH 2: Bush approval vs. exit poll and vote

[kster]

12. kick.nt

[tommcintyre]

13. kick - In tribute to his tremendous contributions - WHY??? :( n/t

[bleever]

14. Kicking for truth, justice, and TIA's invaluable work.