Questioning the Reliability of Univ. of New Hampshire Polls

Over the years, the University of New Hampshire has released some polls that were later proven unreliable, and it appears they are at it again. The college just released the results of their latest study that shows both of the state’s Republican congressmen trailing their Democratic opponents, but the polling methodology is flawed in at least three different ways.

The results show 1st District freshman Rep. Frank Guinta (R) trailing former Rep. Carol Shea-Porter (D), the woman he unseated in 2010. According to the UNH data, Shea-Porter leads the first-term incumbent 44-39 percent. In the more Democratic 2nd District, Rep. Charlie Bass (R) trails 2010 opponent Annie Kuster (D) 39-40 percent.

The first methodological error concerns the length of the sampling period, which stretches from April 9-20. Here, UNH uses a 12-day information gathering period when three is usually considered optimum. The longer sampling time frame has proven to skew results.

Second, the respondent universe of 538 “adults” does not appear to even screen for registered voters. Going beyond the normal voting pool always provides different numbers than one would see from those who are qualified and intend to actually participate in the election.

Third, the number of respondents for each congressional district are low (230 adults in District 1, 251 adults in District 2), thus further deteriorating the reliability factor.

Considering all of the methodology flaws, the current UNH poll should be discounted, though the conclusion that both of the state’s House races will be close and very competitive can certainly be accepted.

Leave a Reply

Your email address will not be published.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