Three Things to Keep in Mind Tonight

Some last minute thoughts before live blogging:

1. Instant polls.  Most of the networks will have their bevy of “uncommitted” voters wired up to give instant feedback on the debate.  Keep in mind the difference between “uncommitted” and “independent” voters – as we discussed in section today, they are not the same.  We know that “Uncommitted” voters who identify with a party are, in the end, likely to support that party’s presidential candidate, and evaluate their debate performance more positively.  They are not “truly” independent voters.  It is the independent voters that Palin must reach tonight if this debate is to change the underlying dynamics of this race.

2.  In a similar vein, be wary of media attempts to measure who “won” the debate using overnight polls of debate viewers, particularly if the polling firm does not provide a breakdown of the partisan composition of the viewing audience.  Recall that the CNN overnight poll that initially suggested Obama won the first debate oversampled Democrats.  Of the debate-watchers questioned in this poll, 41 percent of the respondents identified themselves as Democrats, 27 percent as Republicans and 30 percent as independents. As of the end of August, however, the actual partisan breakdown in the population as a whole was about 38% Democrats, 33% Republicans, and 28% independents. So, Republicans were undersampled by about 6%, Democrats oversampled by about 3% and independents oversampled by about 2%.  It’s not surprising, then, that the CNN poll of viewers judged Obama the “winner” by 13%, 51-38 in the poll.  To their credit, CNN acknowledged the sample bias and in their reporting were clear to state that, given the lopsided sampling, the debate was likely a draw. Not every network is so transparent with the internals of their polling, however.

3. Finally, the key measure when assessing the impact of the debate is not who “won”, but whether any voters’ minds were changed, and/or whether the truly independent voters were persuaded to back a certain candidate.  Very few pollsters bother to ask this question.

Ok, let the fun begin!

Leave a Reply

Your email address will not be published. Required fields are marked *