Skip to main content

View Diary: Why registered voter screens matter—they've been more accurate (206 comments)

Comment Preferences

  •  Problem is a presidential election year doesn't (0+ / 0-)

    look like those highly likely to vote based on past behavior. In fact, most of the people who you get in a pres cycle are people who only vote every four years, if that.

    Second, the questions listed above, attempting to go beyond the simple "do you plan to vote" question in a quest for accuracy, may end up weighting a voter as non likely when they are absolutely likely to vote.

    Thought given to election. (not much. i watched the debate)
    Know where people vote. (ummm same place as last time i hope. it moves around.)

    Voted in precinct before (whats a precinct?)

    How often vote (every 4 years. is that "always")

    Plan to vote (YES)

    LIkelyhood of voting (3 if the lines are too long after work, i wont bother. my guy is gonna win my state anyway *since most voters live in noncompetitive states)

    Voted in last presidential election (YES)

    Is this person a likely voter or an unlikely voter?

    Who the hell knows. It's magic alchemy.

    Better to just go with registered voters and past history. And in presidential election years, the registered voter number is close to accurate because 90% of RVs vote.

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site