Mark Blumenthal | October 8, 2008
A quick follow-up to last night's post on the initial quick reaction debate polls. One point I've tried to make on this subject is that debate usually reinforce existing preferences. Democrats like the Democrat, Republicans like the Republican, so polls that tell us who "won" can sometimes be misleading if the debate audience is skewed toward the partisans of one side or another.
That scenario was not in play last night. In fact, the five surveys for which I could find data all showed a very consistent result. On the question of "who won" or did better in the debate, Democrats consistently gave higher marks to Obama than Republicans did to McCain and independents who could pick a winner typically chose Obama by large margins.
The surveys cited are the two cited last night, from CNN/ORC and CBS/Knowledge Networks (see my post from the first debate for more details on their methodologies), plus the Media Curves online/text message survey, and two automated telephone surveys of debate viewers conducted by SurveyUSA in California and Washington State, two western time zone states where they could make cold calls at a reasonable evening hour after the debate ended.'
Keep in mind that the CBS survey is the only one that limited its reach to those who had been previously uncommitted, that is they were either totally undecided or who might still change their minds. So the smaller number in that survey willing to rate either candidate a "winner" is not at all surprising.