GA: Walker leads Warnock by 6 Points among Georgia Hispanics

[2016]

https://twitter.com/PollProjectUSA/status/1582783427857682442

[Boobs]

Lol

[Holmes]

We’ll circle back to this.

[Unelectable Bystander]

I don’t buy this for several reasons. Also what’s the point of a Hispanic-only poll in Georgia?

[Penn_Quaker_Girl]

I'm no expert on polling (or anything else for that matter), but isn't n=309 a pretty small sample size for a poll?

[Amenhotep Bakari-Sellers]

Abrams is leading and Walker isn't a 50%

[Person Man]

I'm no expert on polling (or anything else for that matter), but isn't n=309 a pretty small sample size for a poll?

I think that's +/- 10.

[GeorgiaModerate]

I'm no expert on polling (or anything else for that matter), but isn't n=309 a pretty small sample size for a poll?

Yes, the margin of error is +/-5.6% (and that applies to each candidate's share, not the margin between them).

[Eraserhead]

There's not enough hispanics in the state for a sample to mean much.

[Penn_Quaker_Girl]

I'm no expert on polling (or anything else for that matter), but isn't n=309 a pretty small sample size for a poll?

I think that's +/- 10.

Yes, the margin of error is +/-5.6% (and that applies to each candidate's share, not the margin between them).

Oh yes how foolish of me.  Definitely understand both of those answers....

[2016]

Biden beat Trump 62-37 among Hispanics in GA in 2020 (7 % of the Electorate)
https://www.nbcnews.com/politics/2020-elections/georgia-president-results

Abrams beat Kemp in 2018 among Hispanics in GA 62-37 (5 % of the Electorate)
https://edition.cnn.com/election/2018/exit-polls/georgia

If either Kemp or Walker are over 40 % among Hispanics in Georgia they are going to win the State.

[GeorgiaModerate]

There's not enough hispanics in the state for a sample to mean much.

Hispanics are about 10% of the state's population.

[RussFeingoldWasRobbed]

Biden beat Trump 62-37 among Hispanics in GA in 2020 (7 % of the Electorate)
https://www.nbcnews.com/politics/2020-elections/georgia-president-results

Abrams beat Kemp in 2018 among Hispanics in GA 62-37 (5 % of the Electorate)
https://edition.cnn.com/election/2018/exit-polls/georgia

If either Kemp or Walker are over 40 % among Hispanics in Georgia they are going to win the State.

Well see about that

[wbrocks67]

Biden beat Trump 62-37 among Hispanics in GA in 2020 (7 % of the Electorate)
https://www.nbcnews.com/politics/2020-elections/georgia-president-results

Abrams beat Kemp in 2018 among Hispanics in GA 62-37 (5 % of the Electorate)
https://edition.cnn.com/election/2018/exit-polls/georgia

If either Kemp or Walker are over 40 % among Hispanics in Georgia they are going to win the State.

Given that there was no change in Hispanic vote from 2018 to 2020, like some other states, then that would make these numbers a bit suspect.

[Person Man]

This election will tell us if Immigration Reform/The Wall debate is becoming more trouble than it's worth.

[RussFeingoldWasRobbed]

Biden beat Trump 62-37 among Hispanics in GA in 2020 (7 % of the Electorate)
https://www.nbcnews.com/politics/2020-elections/georgia-president-results

Abrams beat Kemp in 2018 among Hispanics in GA 62-37 (5 % of the Electorate)
https://edition.cnn.com/election/2018/exit-polls/georgia

If either Kemp or Walker are over 40 % among Hispanics in Georgia they are going to win the State.

Given that there was no change in Hispanic vote from 2018 to 2020, like some other states, then that would make these numbers a bit suspect.

It's possible that an increase in Hispanics population overwhelmed any improvement rs had with hispanics

[GeorgiaModerate]

I'm no expert on polling (or anything else for that matter), but isn't n=309 a pretty small sample size for a poll?

I think that's +/- 10.

Yes, the margin of error is +/-5.6% (and that applies to each candidate's share, not the margin between them).

Oh yes how foolish of me.  Definitely understand both of those answers....

OK, to clarify: the margin of error decreases as the sample size increases (makes sense; a bigger sample gives you a more representative picture of the population).  Computing this for a sample size of 309 yields a margin of error of 5.6%.  This is for a 95% confidence interval, meaning that the actual result will be within (x-5.6%,x+5.6) 95% of the time (there are other types of errors that can creep in, but we'll ignore those). 

So with a result of Walker 47, Warnock 41 and this MoE, we can say that there's a 95% chance that Walker is between 41.4 and 52.6, and that Warnock is between 35.4 and 46.6.   There's a considerable overlap between those two ranges, so it's entirely plausible for Warnock to be ahead with those numbers (and TBF, it's equally plausible for Walker to have a much larger lead than 47-41).

[Penn_Quaker_Girl]

I'm no expert on polling (or anything else for that matter), but isn't n=309 a pretty small sample size for a poll?

I think that's +/- 10.

Yes, the margin of error is +/-5.6% (and that applies to each candidate's share, not the margin between them).

Oh yes how foolish of me.  Definitely understand both of those answers....

OK, to clarify: the margin of error decreases as the sample size increases (makes sense; a bigger sample gives you a more representative picture of the population).  Computing this for a sample size of 309 yields a margin of error of 5.6%.  This is for a 95% confidence interval, meaning that the actual result will be within (x-5.6%,x+5.6) 95% of the time (there are other types of errors that can creep in, but we'll ignore those). 

So with a result of Walker 47, Warnock 41 and this MoE, we can say that there's a 95% chance that Walker is between 41.4 and 52.6, and that Warnock is between 35.4 and 46.6.   There's a considerable overlap between those two ranges, so it's entirely plausible for Warnock to be ahead with those numbers (and TBF, it's equally plausible for Walker to have a much larger lead than 47-41).

Don't you mod-explain to me, I took stat

Seriously though, ah okay that makes sense. 

[Gracile]

There's not enough hispanics in the state for a sample to mean much.

Hispanics are about 10% of the state's population.

A large chunk of GA's Hispanic population are non-citizens or people too young to vote, though. It's estimated that Hispanics make up about 5% of the eligible voting population, and probably a bit less of the actual electorate due to lower turnout rates than black and white voters.

[President Johnson]

All while Abrams leads among them? Throw it in the trash.

[Senator Incitatus]

Any sense of where Georgia Hispanic voters are from? This wouldn't be too crazy if they're just like third-generation Cuban-Floridians who moved up to Atlanta.

[GeorgiaModerate]

Any sense of where Georgia Hispanic voters are from? This wouldn't be too crazy if they're just like third-generation Cuban-Floridians who moved up to Atlanta.

I believe the large majority are Mexican.

[TheDeadFlagBlues]

Do these guy have a track record in polling Hispanic voters?

Edit: "The survey was administered by the School of Public and International Affairs Survey Research Center at the University of Georgia. Surveys were conducted in English." Throw this into the trash - the sample won't be representative of Hispanics. On top of this, a phone poll of Hispanics will be very biased because the challenge of surveying young people via phone is much more important for Hispanics than for the general population. Junk it!

On top of this, only 47% support DACA and 30% of self-identified Liberals want Republicans to win. If this poll is accurate, Democrats are doomed. My guess is that it was conducted by moron undergraduates.

[Unbeatable Titan Susan Collins]

Hispanic Abrams-Walker voters??? One of the weirder split tickets I've heard off.

[GALeftist]

OK