(Note: see Doyle Survey Results link above for a comparison of the Shelburne results to the full poll results)
The other day, I sat down with the Senator Doyle’s Town Meeting Day surveys and decided to “have at it.” What was on the mind of Shelburne voters and how did these surveys compare to the calls and emails I receive? I counted each survey, all 201 of them, recording results, one question at a time. I read longer responses on the back of the forms, then I entered the data onto an Excel spreadsheet and looked at the results.
Let us be very clear here. The Doyle Survey is not a scientific poll. It is not set up to ensure a representative sample has been taken and the system for distribution is not standardized. Since I tabulated the results myself without oversight, citizens must rely on my accuracy and integrity rather than standard survey protocol.
But the survey is not without merit. As WPTZ noted last year, “The Doyle survey is not a scientific poll, but a sampling of views of voters taking part in Town Meeting Day.” After a quick call to my math genius friend John Huntington Smith, I learned that the survey may not be statistically significant but was useful in showing trends. He said the survey has value, “you get a feel, it is not precise, but it is a feel.” We talked about the inherent problems in the phrasing of questions, particularly question # 7. He suggested that instead of showing the results in percentages, I should tabulate them by raw numbers (note: this was presented as numbers in the Shelburne News article, however I recorded them in percentages here to compare with State results).
The danger, of course, comes in interpreting the data, but I couldn’t resist making comments on some of them and have converted those to percentages for easier understanding. Please keep John Smith’s warning in mind.
Survey Trends:
Survey takers overwhelmingly (91%) felt statewide broadband and cell services were important to Vermont and 89% supported mandatory seatbelt laws for adults as well as children. A few thought this law was already in effect, so I asked one of my law enforcement contacts. He said that a seatbelt violation is not a primary offense, however you can be ticketed if you are stopped for something else, speeding for example, and are not wearing your seatbelt.
75% felt that the Vermont prison population should be reduced through the use of alternatives for non-violent offenders; while 71% felt drivers should be prohibited from using cell phones while driving. About 5 people suggested limitations on cell phone use rather than an outright ban. 62% agreed with the idea of a gas tax with a few people suggesting a limit of 5 cents. 55% thought Vermont was not an affordable place to live and 52% of responders opposed lowering the drinking age to 18.
Our toughest questions:
The most controversial questions in our community, based on the press and calls and emails to me, were the success of our schools, same-sex marriage, re-licensing of Vermont Yankee and the performance of elected officials. Half of the respondents thought schools were doing a good job while 30% did not. The “no’s” often noted the time-eternal problem of cost.
Regarding same-sex marriage, of the 201 respondents, 110 supported same-sex marriage while 69 did not. The remaining 22 were undecided.
Renewing Vermont Yankee’s license found 41% of respondents for re-licensing, 39% against with the remainder undecided. One respondent who was unsure added, “Vermont Yankee needs to be studied for efficiency, cost effectiveness and safety. This decision should not be influenced by popular opinion.”
Legislature and Governor
I was curious whether responses to the performance of the Governor and Legislature were related so I scored these questions in relation to each other and sent the data to John Smith. After a dizzying conversation with John about the difference between non-parametric and parametric tests, quartiles, chi-square and rejecting the null hypothesis, John was able to tell me that there was a very high inverse relationship between approval of job performance for the Legislature and the Governor. A “yes” to the Governor’s performance tended to produce a “no” for the Legislature and visa versa.
John thought these numbers suggested that survey takers “had a clear idea of what they wanted to say and it was consistent.” He also thought this reflected the seriousness with which people completed the survey but also warned again that “this should not be generalized beyond the self-selected sample of Shelburne voters.”