Everything about this poll screams spin. There were only 395 responses, and with fully a third coming back as undecided, not voting or won't say, that means only about 260 valid responses. There just aren't enough responses there to trust a weighted average. Especially if the random sample underrepresents a key voting demographic - which the leans table indicates is the case. Carter tries to pass Clark off as a solid second place (20.3%, -3.2 points), but when they remove the weighting, he plummets to a distant third (11.8%, -9.6). That big of a swing indicates the sampling is off somewhere.
Lies, damned lies and Stephen Carter's polling methods.
|