In a recent FiveThirtyEight blog
post, Nate Silver shared a preview of his polling-based forecasts for the Iowa caucuses and New Hampshire primary, the official versions of which he plans to release in a few days. The statistical analysis is based on state-by-state polling; Silver writes that although more objective and subjective factors could be taken into account, "building a well-calibrated forecast from the polls is challenging enough given the uncertainties inherent in primary polling, and so that’s what we’ve decided to focus on rather than anything fancier."
The polling-based analysis for Iowa has Newt Gingrich projected to win 25.1% of the vote, followed by Ron Paul, with 20.7%; Mitt Romney (15.6%); Rick Perry (12.3%); and Michele Bachmann (11.4%); and both Rick Santorum and Jon Huntsman Jr., who are projected to garner under 10 percent of the vote. In his blog post, Silver explains the three factors used to calculate the projections: a weighted polling average, which emphasizes more recent polling as being a better predictor of probable primary results; the reallocated average, "which takes undecided voters and allocates them among the viable candidates;" and a process Silver calls momentum, which "looks at the near-term trend in a candidate’s polling and assigns a small bonus or penalty based upon it."
Silver warns, however, that these projections "are
not very precise, because primary polling is not very precise." He tries to quantify this uncertainty, providing both confidence intervals for each candidate's projections and their win probability, "which combines the vote projections and the uncertainty together to estimate how likely each candidate is to win the caucuses." Gingrich leads with an estimated 49.6% probability, followed by Paul, with 28.2%; Romney (10.6%); Perry (5.2%); and Bachmann (4.1%).
According to Silver's analysis, "Although Mr. Gingrich still has the lead, Iowa looks to be fairly wide open with as many as five plausible winners.
0 komentar:
Post a Comment