Question:
If the 2010 election for U.S. SENATOR were being held today and the candidates included (rotated) Arlen Specter/Joe Sestak, the Democrat, and Pat Toomey, the Republican, would you vote for Arlen Specter/Joe Sestak, Pat Toomey, some other candidate, or aren't you sure how you would vote?

29% Toomey, 28% Sestak
Sestak 38%, Specter 36% among likely voters (
Specter 38%, Sestak 29% among registered voters

About this Poll
The survey findings presented in this release are based on the results of interviews conducted May 3-9, 2010. The interviews were conducted at the Center for Opinion Research at Franklin & Marshall College under the direction of the poll’s Director Dr. G. Terry Madonna, Head Methodologist Berwood Yost, and Project Manager Angela Knittle. The data included in this release represent the responses of 1023 adult residents of Pennsylvania, including 861 registered adults (404 Democrats, 321 Republicans, 117 registered as Independent/Other, and 19 who refused to identify party). Telephone numbers for the survey were generated using random digit dialing, and respondents were randomly selected from within each household. Survey results were weighted (age, education, race, region, and gender) using an iterative weighting algorithm.

The sample error for this survey is +/- 3.1 percentage points. The sample error for registered adults is +/- 3.3 percentage points and is slightly higher for registered Democrats (+/- 4.9 percentage points) and registered Republicans (+/- 5.5
2 percentage points). The subsample of Democratic likely voters has a sample error of +/- 7.9 percentage points.

In addition to sampling error, this poll is also subject to other sources of non- sampling error. Generally speaking, two sources of error concern researchers most. Non-response bias is created when selected participants either choose not to participate in the survey or are unavailable for interviewing. Response errors are the product of the question and answer process. Surveys that rely on self-reported behaviors and attitudes are susceptible to biases related to the way respondents process and respond to survey questions.