If this had been accurate, it would be a major shock - a magical 15-point turnaround in two weeks. In the most recent Pennsylvania polls, Hillary has consistently held either a 10 or 11 point lead. I'm not sure that kind of turnaround is even possible.
The film that did the polling is not even a professional polling group, but a consulting firm which does not disclose which client paid for the poll, and their summary makes statements like "One of the reasons of Trump's surge is attributable to increasing support from the women." (This sentence not only presents a highly unlikely claim, but seems to have been written by someone in third grade. Some of the site's paragraphs are out-and-out illiterate, with subject/verb agreement like "The questions was ..." )
Here we go. As I thought, the whole thing is bogus. They contacted 1000 random owners of landlines and had them respond to a robocall (in English only) by pressing a number to answer each question. And even then they claim a massive 7.1% margin of error. I don't know about you, but I know only one person who still has a landline, and I'm an old coot. About half of Americans own no landline, and there is one cellphone per person in America (and that denominator includes newborns, pre-schoolers, and the institutionalized.)
They balanced the responses to the population by gender only, which pretty much assures that they have excluded the proper proportion of young voters (who may not even know what a landline is). By ignoring a Spanish option, they excluded many Latinos. And their own little wheel graph shows that minority voters only represented maybe 2% of the sample (it's "maybe" because they don't present the raw data to support the graph). So, I think we can say that Trump has a five point lead, give or take seven, among grouchy old white people with landlines. Hell, I'm surprised it's not higher.
Monday, August 22, 2016
"Trump Leads Clinton by 5+ Points in the Key Battleground State of Pennsylvania"
UPDATE: CBS removes report about Trump's lead in Pennsylvania.