Аннотация:The RAND Continuous 2012 Presidential Election Poll (CPEP) was conducted within the American Life Panel, an internet panel recruited through traditional probability sampling to ensure representativeness. Because it asks the same respondents repeatedly about their voting preferences, observed changes are due to individuals changing their minds and not to random sampling fluctuations. The CPEP asks respondents to state both their preferences for a candidate and the likelihood that they will vote as a percent chance. Moreover, we asked the respondents' actual voting after the election, so we can study the predictive power both within sample and out of sample (the national results). The CPEP appears to have predicted well. Our final prediction of the difference in popular vote between Obama and Romney differed about half a percentage point from the final tally, which would place it near or at the top of the polling firms. The probabilistic questions, even months before the election, were strongly related to actual voting behavior. Our approach allows us to gain insights in stability of voting preferences and the effect of events on individual preferences; for example, various shifts can be clearly related to major events. The American Life Panel has a wealth of background characteristics that can be related to voting preferences.
