- Joined
- Oct 31, 2004
- Messages
- 32,748
- Likes
- 39,376
Difference in sampling technique (RV vs LV).
Just for fun on the credibility scale; CNN recently ran stories claiming the lowest approval scores for Trump since he was president. When you looked behind the numbers they cited raw data findings rather the the RV sample they typically post (and you see in your link). The RV number in their poll was 5 points more favorable to Trump but they found a way to analyze it in the worst possible way to create a story (their reported low number was much further out of line than Rasmussen is in your link).
Bottomline, don't trust any single poll and changing who you include in your sample changes your number. The LV approach argument has merit.
I'm aware.
You may have missed the strutting around last week when Rasmussen had them even and suddenly it was a credible poll and no longer fake news.