In a very worrisome email, a politically involved FrontBurnervian challenges the News’ mayoral poll, as well as the Bradley effect. This is dynamite stuff, so please read carefully. Because if the News did tamper with the poll results, as our FBer suggests — either because it was being accused of a pro-Leppert bias or for other reasons — the only conclusion to be reached is that it is no longer worthy of trust. I hope someone in the News management or with the polling organization with respond with specifics.
Maybe the Bradley effect exists, but I don’t think it’s applicable at all here … It’s more likely that the Morning News and/or their pollster manipulated the numbers to get a closer race. You have to wonder why they would sample 900 people, then publish a poll with a 280 person sample. I’ve been in this business a long time, and I have yet to meet a pollster who will stand by sample that is so small … Not statistically sound. Also, the Morning News poll was 21% African American. Why would they weight Af Am so high when the Af Am turnout on May 12 was only 16% (with Don Hill on the ballot)? They probably did this because it would give them a closer race and they could justify over-sampling AfAm since there were run-offs in South Dallas. But truth is, the likelihood of getting a AfAm turnout of 21% was not likely.