First, it's data were, not was. Second, it's confidence level, not confident level. Now, the irony. This is a student organization, yet it used what appears to be a robo-poll, which by definition cannot call cell phones -- the very key aspect of college students' very existence.Data was collected onOctober 25-30 using an automated data collections system. The Virginia sample consisted of 874 registered, likely voters with a margin of error of +/-3.24% at a 95% confident level.
Plus, robo-polls -- those annoying kind over the phone where it's not a real person -- also tend to skew older, Republican, conservative. Hence, the idea that in this poll the gap is smaller (favoring the GOP candidate in this case) raises concern, especially compared to a different poll that had a wider gap.