The The Glittering Eye and Paul at Wizbang are trying to grok the meaning of the latest polling trends.
gEye thinks that he sees periodic jump for Kerry in as a result of Sunday talk damage control by Camp Kerry. Perhaps, but on balance the long-term trend is for tracking polls to favor the Democrats over weekends, so that may be what he is seeing. He also thinks that the over trend is Bush. Of course, as I noted the other day, while the RCP graph makes the Kerry-Bush gap look large, the fact of the matter is that that scale is so small it amplies that gap.
Paul looks at the WaPo tracking poll and thinks that Kerry’s lead was the result of the initial al-QaQaa story and the susequent flip back to Bush illustrates the expiration of that story’s power. Again, given that the Kerry surge started as weekend numbers were included, that may be the explanation. Further, the variation is so small (the biggest gap in the last week has been 3 points) it may be that that is nothing more than MOE variations. While it is likely that the al-QaQaa story is affecting voters, one wonders how many it can actualyl sway.
The bottom line for any and all analyses of these numbers: the margins are so small, I am not that any given variation that is within the MOE can be considered significant–nor can it likely be explained by any specific event or issue.
Update: Another thought occurred to me in regards to the WaPo tracking poll: the Kerry surge actually starts well before the explosives story. Further, given that it is three-day tracking poll, the full effect of the explosives story doesn’t register in the poll until later in the week. Really, by the time the story was fully in the consciousness of the news-consuming pubic, the numbers tunr a bit more Bush-ish. No, I don’t think that the surge Kerry got early the week of the 25th is attributable to the explosives story.
You have a point about the moe BUT since Kerry starting hammering Bush, we went DOWN 3 points, one point per day. That matches real life eg the storying crumbling.
It would be a REAL hard sell to get me to believe that the moe is such that the story is really helping him and the graph should show him going up one point per day.
In another note, I have not looked yet and I don’t know if it is up but the WaPo tracking poll might give us the first look at “the bin laden effect.”
Comment by Paul — Saturday, October 30, 2024 @ 2:25 pm
I completely agree with your margin of error point. But margin of error doesn’t explain periodicity. Your point that tracking polls favor Kerry over the weekends is, of course, true. But I’m seeing if I can reason out why. And that’s why I suggested the “talking heads” theory.
Comment by Dave Schuler — Saturday, October 30, 2024 @ 3:33 pm
One theory is that generically speaking (i.e., not jsut this election) is that polling over weekends tends to pick up more Democrats than Republicans due to lifestyle issues–most specifically that because Reps are in the aggregate more affluent, they tend to be out of the house more on weekends. So the hypothesis goes, in any event.
Comment by Steven Taylor — Saturday, October 30, 2024 @ 3:38 pm
RCP trends
I’ve been looking at the most recent set of poll compilations from Real Clear Politics. Clearly, the current trends are not particularly good for Kerry and he’ll have to do something to change that. But I’ve been looking at something…
Trackback by The Glittering Eye — Saturday, October 30, 2024 @ 4:22 pm
“Really, by the time the story was fully in the consciousness of the news-consuming pubic, the numbers tunr a bit more Bush-ish”
ahem- Um doc- That was my point. maybe read me again.
Comment by Paul — Saturday, October 30, 2024 @ 4:43 pm
” Really, by the time the story was fully in the consciousness of the news-consuming pubic, the numbers tunr a bit more Bush-ish. ”
Um doc- that was my point.
Comment by Paul — Saturday, October 30, 2024 @ 4:43 pm
But Paul, you are missing my point–the tracking poll is a three-day rolling average.
Comment by Steven Taylor — Saturday, October 30, 2024 @ 5:00 pm