Statistikhengst
Diamond Member
- Banned
- #1

Marist put out a small write-up on why it polls RV now and LV closer to an election:
6/2: Likely Voters: Why Later Not Now | Home of the Marist Poll: Pebbles and Pundits
It's a pretty good, small write up.
I agree with some of it. Marist also goes some into the "why" of early polling.
Go read it and let us know what you think. This should be a topic where people who usually go at each others' necks should be able to have a civil and adult discussion.
-----------------------------------------------------------------------
BTW, in both 2008 and 2012, the AVERAGES of polls (usually a 14 day time-frame, 7 days right before the election), using both RV and LV values, were very, very good. The values were better in 2012, they were somewhat more slanted to the Right in 2012. This has taught me that the "truth" about an electorate usually lies somewhere in the middle between these two possible calculation models. So, I neither praise nor do I damn either of these models.
Last edited: