The poll, conducted April 30 and May 1 by Republican pollster Christine Matthews of Bellwether Research and Fred Yang of Garin-Hart-Yang Research Group, surveyed 700 likely voters in next Tuesday's primary. The margin of error is plus-or-minus 3.7 percentage points. All respondents have a history of voting in Republican primaries, the pollsters reported, with slightly more than three-quarters of respondents identifying as Republicans. Matthews writes that the GOP Senate primary could "attract some voters with no history of voting in a Republican primary." To account for that, pollsters "also dialed a separate list of voters with no primary vote history but who voted in the 2008 and 2010 general elections," she writes, but did not call those who had voted in Democratic primaries. Those calls were discontinued after the first night for two reasons: First, Matthews writes, the calls "are expensive" since few of these voters said they were likely to vote in the GOP primary, and, second, they broke for Mourdock "in a similar fashion to our ... sample of past GOP primary voters." Since Mourdock has a significant lead in the overall poll, an expanded electorate is unlikely to be a decisive factor in the race; as a result, Matthews writes, the 50 interviews conducted on April 30 with voters with no past GOP primary vote history were ultimately not included.