NightTrain
03-06-2016, 09:49 AM
I was looking into the Florida race this morning.
All of them are consistent, with Trump at 45% and Rubio at 25%.
But then there's one outlier poll saying Trump is 35% and Rubio at 30%... this was conducted by the Tarrance Group, who the Miami Herald identified as an Anti-Trump group.
http://heavy.com/news/2016/03/florida-gop-republican-polls-latest-current-polling-numbers-favorite-primary-donald-trump-marco-rubio-ted-cruz/
I get that people play up their guy and downplay the other guy... but if you're doing a poll, what kind of logic is there to promote bogus numbers? Is it a case of presenting numbers to appeal to groupthink? Or is it a case of giving the appearance of a close race in order to urge people to vote?
In any case, in the end, the only thing that can be certain by doctoring the numbers and presenting a skewed result is that you aren't a reliable source - but I suppose that's not a consideration if you're using a throwaway name. I've never heard of the 'Tarrance Group' before, but they've been around for a couple years and worked for Kasich - not that that's any kind of crowning achievement.
I remember when I was living in Fairbanks one election season in '99, a robot from the University of Alaska Fairbanks called the house and began asking me questions about something locally related, and the questions were structured to promote an Uber-Greenie kind of response. I answered honestly, and anyone that knows me knows that my answers were decidedly against what such a pollster would want.
The robot hung up on me halfway through the poll, literally as it was asking me another question. 2 days later there was a story in the local paper about how an overwhelming majority of locals supported this new Greenie initiative. I'm quite sure my responses were scrubbed from the results, as well as the majority of Alaskans who answered like I did - not many Alaskans subscribe to liberal / greenie notions. But they got the results they wanted, come hell or high water.
That was a case of someone trying to get their agenda passed, I understand completely how that worked. I suspect that disregarding respondents answers that you don't like is the same here with Tarrance - filter it out until you have what you want.
But skewing the poll results when the truth of the matter will be out there shortly? You've only given yourself a black eye in the long run.
All of them are consistent, with Trump at 45% and Rubio at 25%.
But then there's one outlier poll saying Trump is 35% and Rubio at 30%... this was conducted by the Tarrance Group, who the Miami Herald identified as an Anti-Trump group.
http://heavy.com/news/2016/03/florida-gop-republican-polls-latest-current-polling-numbers-favorite-primary-donald-trump-marco-rubio-ted-cruz/
I get that people play up their guy and downplay the other guy... but if you're doing a poll, what kind of logic is there to promote bogus numbers? Is it a case of presenting numbers to appeal to groupthink? Or is it a case of giving the appearance of a close race in order to urge people to vote?
In any case, in the end, the only thing that can be certain by doctoring the numbers and presenting a skewed result is that you aren't a reliable source - but I suppose that's not a consideration if you're using a throwaway name. I've never heard of the 'Tarrance Group' before, but they've been around for a couple years and worked for Kasich - not that that's any kind of crowning achievement.
I remember when I was living in Fairbanks one election season in '99, a robot from the University of Alaska Fairbanks called the house and began asking me questions about something locally related, and the questions were structured to promote an Uber-Greenie kind of response. I answered honestly, and anyone that knows me knows that my answers were decidedly against what such a pollster would want.
The robot hung up on me halfway through the poll, literally as it was asking me another question. 2 days later there was a story in the local paper about how an overwhelming majority of locals supported this new Greenie initiative. I'm quite sure my responses were scrubbed from the results, as well as the majority of Alaskans who answered like I did - not many Alaskans subscribe to liberal / greenie notions. But they got the results they wanted, come hell or high water.
That was a case of someone trying to get their agenda passed, I understand completely how that worked. I suspect that disregarding respondents answers that you don't like is the same here with Tarrance - filter it out until you have what you want.
But skewing the poll results when the truth of the matter will be out there shortly? You've only given yourself a black eye in the long run.