...that the Democratic party infighting helps McCain more than anyone else.

Obama keeps this up and his numbers will look as bad as Clinton.

"Contrary to mainstream belief, it appears that the more people get to know him, the less they like him."

That claim seems to be undermined by the fact that Obama has been winning primaries in states where he used to be down by big numbers to Clinton.

The low democratic numbers only reflect the polarized nature of the primary.  We will come together regardless of the nominee, and trounce McCain come Nov.

THIS was the graph you were crowing about earlier?

I hardly know where to begin mocking :')

My objection to the 1990s is simple: I was young in the 1990s, and youth was wasted on me.

Dear SevenStrings,

Good work on the graph.

My recommendations:

--One month does not seem adequate to track "long term trends." I suggest going back much further. The result will render a larger downfall for Obama, so you'll probably like that. However it will also probably (an educated guess) demonstrate a certain stability in HRC's numbers--especially her negatives.

--Ultimately, "favorable" trends in politics are pretty malleable--just look at Bush's post 911 spike. "Unfavorables" especially strong unfavorable ratings tend to take longer to change. Once you've got a negative opinion of someone it's usually harder to change, right?

--I don't find your principle data factor Postive-negative to be particularly meaningful. To me, the more meaningful number is the raw negatives, and even MORE meaningful are the scaled raw negatives.

Negatives are a clear indicator of how "divisive" a candidate is--especially numbers for how many people STRONGLY DISLIKE a candidate. When the nagative number goes above 50% a candidate needs some deep PR work. HRC, for example, to be electable will need just that kind of PR work. For me, it was the movie "the hunting of the president." As someone who had a generally negative feeling about HRC (because I didn't like what I saw as a combative leadership style) I shifted to a strongly favorable opinion after seeing that movie. My current opinion of her, due to the campaign, borders on hostility (I too, am effected by emotions.)

So I would conclude that if America could get to know Hillary, they would (mostly, you can never count on the freepers) like her, but that this is unlikely in a long campaign that's reinforcing the somewhat sexist opinion of her as "too combative."

--In the GE, both of our candidates will have the "cognitive dissonance" advantage over McBush. Honestly, I think that there are a lot of white folks who wont want to have a negative opinion of Obama because they're afraid that will mean they're racist. I think HRC gets a little of the same, people wont want think that they're voting against her just because she's a woman. For this reason, I think it's too early to make "vs. McCain" predictions from this data.

--I agree that the short duration of data and the large MoE make the graph largely meaningless in terms of the trends you claim it represents. Looking at the raw data, here's what sticks out to me:

1. HRC's negatives are stable around 50.7% or so.
2. Obama's negatives have moved up from an average of about 44% to about 46%. This is the effect of a very negative campaign cycle, and yet his negatives have not broken 50%. Based ob this, I'll  predict that he's pretty resilient, and that his negatives are likely to stay below the unelectable mark and even come down during the GE with McCain (due to the consolidation of Democratic support and the congnative dissonance factor above.)

Finally, I think an actually meaningful graph would extend back before January and would separately chart numbers for negatives and positives with the ghost number of P-N indicated in between. That would allow the reader to draw their own conclusions.

Peace.