Was There Sampling Bias?
Although the polling results summary identifies four different sub-groups within the general sample, most questions were either asked to the general sample, or the sample consisting of Republican primary voters. Questions regarding enthusiasm to vote in the general election were directed to both Republican primary voters and Democratic primary voters. Since President Obama is not being contested for the Democratic nomination, there would likely be more respondents planning to vote in the Republican primary than any Democratic primary, however sizes of the sub-groups were similar: 340 for Republican primary and 367 for the Democratic primary. These may indicate a general sample that overwhelmingly leans Democratic. One of the polling questions asked all respondents whether they are “more likely to vote in a Democratic presidential primary or caucus, a Republican presidential primary or caucus, or aren’t likely to vote in any presidential primary or caucus at all.” If this was the question used to identify the sub-groups, Democrats may have answered that they would be more likely to vote in a Democratic primary than a Republican one, even if they did not plan to actually vote in the presidential primary this election cycle, making the spread between parties more likely to be balanced, but also making the categorization of the sub-groups somewhat misleading.
Analyzing the Poll Results
Despite some lack of clarity regarding question wording and presentation, the CBS/NYT poll mostly met good polling practices, however spreads in aggregate polling data are currently large. Seven GOP primary polls have been conducted since January 11, including the CBS/NYT poll. Of these, only CBS/NYT and Rasmussen have Romney ahead by a single digit spread, seven and three points, respectively. Other pollsters including Gallup, Pew Research, ABC/WaPo, Fox News, and CNN have Romney with a double digit lead, ranging from 16 to 25 points. All polls are fairly consistent with numbers for Santorum and Paul, with ranges between 13 and 16 points for Santorum, and 12 and 16 points for Paul. Gingrich’s numbers are varied, ranging from 14 to 27 points, with higher margins trending in the more recent polls. CBS/NYT and Rasmussen reported the highest percentages for Gingrich.
GOP Polling: Another Gingrich Surge?
This data suggests that a drop in support for Romney, counteracted by a Gingrich surge, may be an emerging trend, or may be due to polling flaws by the two pollsters that are inconsistent with the aggregate trends.
New York Times. Full Results of the New York Times and CBS News Poll. How the Poll was Conducted. Accessed January 19, 2012.
CBS News. Poll: Mitt Romney Atop Fluid GOP Race with 28 Percent. Accessed January 19, 2012.
RealClearPolitics. 2012 Republican Presidential Nomination. Accessed January 19, 2012.
Decoding Science. One article at a time.