If Trump did commission it, probably nobody.
Yes, we decided not to do what all the professional pollsters do in order to “provide a more representative ground reality”. Yeah, ok. But if you are correct and they only had 190 respondents then maybe it was simply too small to weight it by other demographics factors.
They also say:
Given that quote, I wonder what proportion of the poll respondents were women to begin with, and if they simply cherry-picked the one weighting factor that would wind up inflating Trump’s numbers.
Read the comments on The Gateway Pundit. Most of them are loudly shouting that the mainstream media has rigged all the polls. They aren’t questioning this poll or its dubious founders. They’re welcoming it as the Truth finally shining its light on Crooked Hillary and her gang. Looks like the Freepers have found a new home with better web design.
You know what? Let 'em.
If they think there’s some advantage to lying to themselves who am I to dissuade them? They may be thinking that there’s some sort of influential impact of publishing tainted data for electoral gain but they’re deceiving themselves.
That’s my guess. With such a small sample, probably biased, and so many possible covariates, it’d be trivial to build a few dozen models and cherry pick one with any desired outcome…
Once again I would have to ask: why fake polls if you’re down? Personally, being down would motivate me to go to the polls. I don’t want to see my side being up 10 points, I want to see us down so I would get off my lazy ass and go vote. Maybe most Americans are difference, maybe they see themselves up and go vote, and the other side sees themselves down and doesn’t vote, but it never made any sense to me why anyone would bother. Unless we have data showing people are more likely to vote when they are down or up
My suspicion is that the thought is: “if our supporters feel that our campaign is a lost cause (due to being well behind in the polls), they won’t even bother to come out and vote.” Some polls are pretty accurate, some aren’t…but if the actual sentiment is actually closer than the polls suggest, then that idea that “the polls have us way behind” could kill a campaign that might otherwise stand a chance.
The Alberta economy in general, and the energy sector in particular, has been really hurting because of oil prices. So I can’t fault these guys for trying to diversify when their old clients are drying up. “What else is can we do with our knowledge of statistics and is a really hot market right now?”
But I think they must be either really overconfident to release something so far outside the orthodoxy, or really shifty in trying to get into the “unskewed” market.
I don’t know which way it pulls but how close an election seems to be does seem to affect voter turnout
I suspect that this particular poll has no connection to any campaign or group of supporters. It seems more likely they’re doing it for funsies, and deliberately reaching a conclusion opposite that of prevailing legit polls so that people will pay attention.
True to an extent, but in this election we have the spectre of Trump bleating about how the election will be “rigged”, and whipping his supporters up into a frenzy about biased media, and exhorting them to do things about it. This crap poll is already being used by the conspiracy theory nuts to show how the media is hiding the “real truth”. Not good in the long run.
I think you may have hit on the most likely explanation. It’s a way of getting attention for their new business (and as vd said, the Oil and gas business in Alberta is in the dumper, so what else are they gonna do?
Generally speaking, an election perceived to be close results in a higher turnout. Races in which the outcome seems assured lowers turnout. I’m not finding anything in a quick search that says that being up or down in the polls spurs voting. Putting out results that indicate a competitive race is probably the best strategy to get people to vote.
One thing we do know is that people like to vote for the winner. It’s gives them a feeling of victory and connection. That’s why polls after the fact normally show a much higher percentage of respondents claiming to have voted for the winner than the actual percentage.
“the sample was balanced based on the 2014 Pennsylvania Census with weighting factors only for gender” - wait, what? Did the state conduct its own census in 2014? Or are they using American Community survey data from a period ending in 2014, which isn’t actually a census?
Who knows?
These two guys are chemical engineers from Alberta, who seem to think that this background gives them the expertise to do political polling in another country.
Can’t wait to see them produce a report that contradicts the Department of Labor Statistics in the US, and claims that unemployment is really 40% like Trump says. Or maybe they’ll branch off into astrophysics - can’t be that hard, can it?
I’m looking forward to their upcoming white paper on climate change.