由买买提看人间百态

boards

本页内容为未名空间相应帖子的节选和存档,一周内的贴子最多显示50字,超过一周显示500字 访问原贴
USANews版 - 如何正确的作poll
相关主题
某种理论推测trump比你想的支持率更高War over polls intensifies
看看那个州长反超的poll是不是扯淡PA的新Poll,川普领先了
原来Conway还有个polling firm新的FL Poll
538: Election Update: Why Our Model Is More Bullish Than Others On Trump538 网站警告: 早期出口民调错误很大,常常夸大民主党选民
号外号外,新的poll来了!你们这些动不动研究poll的希粉根本没有美国政治常识
Polls are always wrong in close but polarized elections!佐治亚民主党州长候选人:蓝浪也包括 非移!牛大了!
Gallup stays out of 2016 presidential polls看Poll得区分它们的质量
can't trust polls最近新出的几个poll很让人担心啊
相关话题的讨论汇总
话题: cahaly话题: trump话题: trafalgar话题: senate话题: florida
进入USANews版参与讨论
1 (共1页)
d********f
发帖数: 43471
1
Pollster Who Got It Right in 2016 Does It Again
Robert Cahaly, senior strategist for the Trafalgar Group, made a name for
himself in 2016 by being the only pollster to correctly show Donald Trump
with a lead in Michigan and Pennsylvania -- two key states he carried --
heading into Election Day. (He did not poll Wisconsin, another surprising
win for Trump.) Cahaly also showed Trump ahead in North Carolina and Florida
, both of which he won, securing his improbable 304-227 Electoral College
victory over Hillary Clinton.
Cahaly managed to pick up support for Trump that all other pollsters missed
by employing a unique method that sought to measure support from voters who'
d been “inactive” in recent election cycles, as well as adding a question
to his surveys designed to isolate the effect of social desirability bias
among Trump voters – the concept that people won’t tell pollsters their
true intentions for fear of being stigmatized or being politically incorrect.
After asking voters who they were supporting in 2016, the pollster followed
up by asking them who they thought their neighbors were supporting, Trump or
Clinton. Cahaly consistently found a high degree of variance between who
respondents said they were voting for and who they thought their neighbors
were voting for, suggesting there was in fact a “shy Trump effect” at play.
Two years later, Cahaly’s method once again proved solid. In one of the
most polled races of the cycle, Trafalgar stood alone as the only polling
firm to correctly show a Ron DeSantis gubernatorial victory in Florida – as
well as Rick Scott winning the Senate race there. (Both narrow outcomes
will likely result in recounts.)
Trafalgar also correctly predicted Senate outcomes in Missouri, Montana,
North Dakota, Texas and West Virginia (as of this writing Arizona remains
undecided), making it the most accurate pollster of the cycle among those
firms that polled multiple Senate and governor races.
Even with Trump not technically on the ballot, Cahaly found social
desirability bias playing a role. In the Florida and Georgia governor
contests, some of that was attributable to race – often referred to as “
the Bradley effect” – but Cahaly also found a “shy Trump” effect playing
a role in places like the Arizona Senate race.
In the Georgia governor race, Cahaly’s results showed a big win for
Republican Brian Kemp over Democrat Stacey Abrams. The final result, however
, was a much closer Kemp victory. Cahaly gave credit to the Abrams campaign.
“They did a great job of registering voters late and bringing lots of new
people into the process,” he told RealClearPolitics by phone on Thursday.
The same was true in Texas, Cahaly said, where Beto O’Rourke finished just
2.6 percentage points behind Ted Cruz.
Still, the pollster believes his method and methodology will be more
valuable than ever as low voter response rates and social desirability bias
continue to present challenges to all pollsters in the future.
“What few mistakes we made this time, we won’t make again,” he said.
Based on its record in 2016 and this year, Trafalgar is a polling firm to
watch as Trump’s 2020 re-election bid heats up -- and an expected large
field of Democrats jockey for the right to oppose him.
d********f
发帖数: 43471
2
poll的问题设计很重要,如果让人说真话,如果你一边问你是不是racist,一边你问你
是不是trump supporter那你觉得你能问出啥来

Florida
missed
who'

【在 d********f 的大作中提到】
: Pollster Who Got It Right in 2016 Does It Again
: Robert Cahaly, senior strategist for the Trafalgar Group, made a name for
: himself in 2016 by being the only pollster to correctly show Donald Trump
: with a lead in Michigan and Pennsylvania -- two key states he carried --
: heading into Election Day. (He did not poll Wisconsin, another surprising
: win for Trump.) Cahaly also showed Trump ahead in North Carolina and Florida
: , both of which he won, securing his improbable 304-227 Electoral College
: victory over Hillary Clinton.
: Cahaly managed to pick up support for Trump that all other pollsters missed
: by employing a unique method that sought to measure support from voters who'

d********f
发帖数: 43471
3
Two years later, Cahaly’s method once again proved solid. In one of the
most polled races of the cycle, Trafalgar stood alone as the only polling
firm to correctly show a Ron DeSantis gubernatorial victory in Florida – as
well as Rick Scott winning the Senate race there. (Both narrow outcomes
will likely result in recounts.)

【在 d********f 的大作中提到】
: poll的问题设计很重要,如果让人说真话,如果你一边问你是不是racist,一边你问你
: 是不是trump supporter那你觉得你能问出啥来
:
: Florida
: missed
: who'

a****1
发帖数: 634
4
这算新发现!!??老天,我都在本版说过:投共和党的人在民调时不太愿意讲真话。
我也说过,早年台湾支持民进党的人也不太愿意说真话。这道理非常简单,用中文讲叫
做不好意思,主楼的那位发明的名词 social desirability bias 是同一个意思。这叫
做隐性民意。
当然,这个现象尽人皆知,一点不稀奇,但主楼那位的贡献可能在于如何设计问题,把
这些隐性民意找出来吧?
r******g
发帖数: 4002
5
白等怎么有可能当选总统?如果大选日,他的名字还在选票上,对我来说就是一个意外
1 (共1页)
进入USANews版参与讨论
相关主题
IN: TRUMP 37 CRUZ 35号外号外,新的poll来了!
现在的polls和初选的不一样 不要太在意Polls are always wrong in close but polarized elections!
Reuters:希拉里民调不行,我们改也要改得她行!Gallup stays out of 2016 presidential polls
说说NBC/WSJ的Pollcan't trust polls
某种理论推测trump比你想的支持率更高War over polls intensifies
看看那个州长反超的poll是不是扯淡PA的新Poll,川普领先了
原来Conway还有个polling firm新的FL Poll
538: Election Update: Why Our Model Is More Bullish Than Others On Trump538 网站警告: 早期出口民调错误很大,常常夸大民主党选民
相关话题的讨论汇总
话题: cahaly话题: trump话题: trafalgar话题: senate话题: florida