Another Poll, Another Chafee Lead (but smaller)
New Hampshire polling outfit American Research Group (via RIFuture) seems to have done a more robust job of polling on the RI GOP Primary race than some recent efforts. The poll methodology was “based on 384 completed telephone interviews among a random sample of likely Republican primary voters in Rhode Island from April 25-May 2, 2006. The theoretical margin of error is plus or minus 5 percentage points, 95% of the time.” The poll question was:
If the primary election for US Senate were being held today between Lincoln Chafee and Steve Laffey, for whom would you vote – Chafee or Laffey? (names rotated)
And the results broke down as follows:
Likely primary voters — 48% Chafee / 39% Laffey / 13% Undecided
Rep. (56% of sample) — 46% Chafee / 42% Laffey / 12% Undecided
Ind. (44% of sample) — 51% Chafee / 35% Laffey / 14% Undecided
ARG also has a neat thing called the “Ballot Lead Calculator,” which tries to “take into consideration the sampling error for the difference between two estimates that are derived from the same sample.” I entered the numbers as instructed, and it revealed that Sen. Chafee has an upper margin of victory of 18.3% and a lower of -0.3%. This means, according to ARG, that “there is no significant difference between the ballot numbers for candidates A [Chafee] and B [Laffey] – they are statistically tied in that poll.” This is probably largely because of the significant number of undecideds.
Now, all of you amateur pollsters can play around with what you think the breakdown of Republicans versus Independents will really be in the GOP primary and come up with whatever number satisfies you!