(This is a BIG graphic. Click once or twice for full resolution.)
This post tracks the national Republican vs. Democrat trial heats for the 2008 presidential election. It is updated in place so a link to this page will always come to the most recently updated trial heat data.
Warning note: Recent polling has given most attention to pairings of Giuliani or Thompson against Clinton or Obama. Romney and especially McCain have been less likely to be included in a test. Likewise Edwards has often been omitted from trail heat questions. This means that the current estimate for McCain vs. Edwards, for example, is heavily influenced by the single recent poll that tested that pairing. When using these charts, look to see that the latest estimates are based on several recent polls. If not, then the results will primarily reflect just the single most recent result. That does not mean the trend estimate is wrong, just that it is based on less data than we usually would like to have.
There is a link to this page from the Trial Heat thumbnail in the right-hand column.
The trend line is only estimated and plotted when there are at least 8 data points to fit. If there are fewer, then only the points are plotted. The estimated candidate support in the figure heading is the latest trend estimate if there are 8 or more polls. Otherwise it is the median of the polls that are available.
The "latest poll" can vary across pairs of candidates because not all polls ask about all pairs.
Polling has avoided trial heats among the least known candidates so far. More will be added as data become available.
I'll avoid commentary here, where the point is just the display of the data. For commentary and other posts on the 2008 Presidential race, click on the "Pres08" label below or on the link in the right column of the page under "2008 Trial Heats".
