If you've been reading this blog you've known for more than a month now that the Reuters/Ipsos poll has severe methodology issues, consistently over-sampling Democrat voters by as much as 10% of the overall sample. As a result of this consistent, unaddressed issue, the poll showed Hillary Clinton leading by 15% one day in mid-July when no other poll had her lead registering at over 5%, with most having it within 1-2%.
Even though this ridiculous sampling method produced wildly-outlying results, Reuters/Ipsos did nothing to change its method until last week, when, even with its over-sampling of Democrat voters, the poll showed Trump suddenly moving into the lead following the GOP National Convention.
Now suddenly the folks who run the poll are recognizing they have a systemic problem in their polling method and are moving to "fix" it - only they're NOT fixing the problem that really exists, just "fixing" the method so that it will reliably show Hillary back in the lead.
Your fake national news media in action, folks.
No comments:
Post a Comment