The Republican Primary Race Is Filling Up

[MisterChester]

I think you're on the right track. I believe this will indeed be a historic election and witnessing a bit of it live would be memorable. Unfortunately I've seen enough of Kasich to last a lifetime

Seen enough of Kasich eating, you mean?

 

[BugI02]

Worse than that. I live in the Columbus 'burbs. My lifetime exposure is beyond safe limits

 

[jamil]

In my opinion, "conservatism" is an ideology that regards the degree to which individual liberty and state control interact/intersect. The ideology lies on a spectrum, at the opposite end of which is "statism". Conservatism, IMHO, is articulated most eloquently in the Declaration of Independence, and favors individual liberty over state control, while recognizing that certain, limited state is required to ensure that individuals can maintain a civil society.

Undoubtedly, conservatism itself is a spectrum - which is why I left the definition as vague as I did, articulating principles rather than specifics. And under my definition of "conservatism", using the government to legislate morality would fall outside of that spectrum, and would instead be part of the statist spectrum.

As for shifting definitions: that is true, and known. In fact, "liberal" and "conservative" have almost completely exchanged definitions over the past half-century.

I forget what it's called when you have more abstract words that we use to refer to something more concrete, and then the abstract word because the common usage. "Liberal/progressive" and "conservative" are relative terms. The concrete thing we're referring to depends on from what we want to progress and what we want to conserve. If we're going to use one dimension I think a telling dichotomy is individualist vs collectivist. But that would mean some people change sides from the traditional liberal/conservative groupings.

Abandon all hope. Sanders outpolling Cruz and Trump both by double digit margins and beating Kasich by almost 5.

And even Sanders can't beat Clinton






Or maybe something is wrong with the polls

Yeah, there's something wrong with the polls. It's a real world example of how non-transitive polls have conflicting outcomes. If you want to know who really is the candidate people like the best, you don't really get then when candidate A > candidate B > candidate C > candidate A. But that's exactly what we have.

If Clinton beats Sanders, and Sanders beats Trump, and Trump beats Cruz, and Cruz beats Kasich, then Clinton should certainly beat Kasich. But polls consistently show that Kasich outpolls Clinton. It's because the pols don't capture how each candidate stacks up against all the candidates head to head. The poll question just asks about two candidates in isolation from the others.

If their polls asked voters to rank the candidates, it the outcome of that would be transitive. A > B > C, and C < A and would be a more accurate representation.

 

[BugI02]

I also think since poll subjects to a certain extent self-select (by being willing to answer the phone/text/tweet/whatever form of intrusion) that perhaps a 'representative sample' may obey the correct mathematical formula but become less and less representative.

But I kind of like the cliff-hanger, outcome-in-doubt nature of the remaining primaries and the coming election. It manages to hold my interest like few other events

 

[Timjoebillybob]

I also think since poll subjects to a certain extent self-select (by being willing to answer the phone/text/tweet/whatever form of intrusion) that perhaps a 'representative sample' may obey the correct mathematical formula but become less and less representative.

They have formulas that factor that into account. It's part of the margin of error. There was a poll posted on here recently that put Trump up close to or possibly beating Hillary in winning the general. But it had over a 14 point MoE. In large part due to how the participants were selected. The actual polling was done by survey monkey which has a pool of those that sign up to take online surveys/polls.

 

[GodFearinGunTotin]

Donald Trump chose to brand his foreign policy with the noxious slogan "America First," the name of the isolationist, defeatist, anti-Semitic national organization that urged the United States to appease Adolf Hitler.

Trump's America First has ugly echoes from U.S. history - CNN.com

Because nobody on the Trump campaign knows how to use Google? Yeah, that's it...that gotta be it..........doesn't it?

 

[rob63]

Trump's America First has ugly echoes from U.S. history - CNN.com

Because nobody on the Trump campaign knows how to use Google? Yeah, that's it...that gotta be it..........doesn't it?

The irony being that isolationist, defeatist, anti-Semitic appeasement of totalitarians doesn't sound like too great of a change from the current policy.

 

[T.Lex]

New Indiana poll:
http://cloutpolitical.com/wp-conten...GOP-PRIMARY-POLL-TOPLINE-REPORT-4-28-2016.pdf

Seems like pretty good methodology, with even spread among cited demographics: Trump 37, Cruz 35, Kasich 15, MOE ~5. So, the leaders are within the MOE.

Poll was 2 days ago, so the Fiorina thing was still hot, I think.

 

[chipbennett]

New Indiana poll:
http://cloutpolitical.com/wp-conten...GOP-PRIMARY-POLL-TOPLINE-REPORT-4-28-2016.pdf

Seems like pretty good methodology, with even spread among cited demographics: Trump 37, Cruz 35, Kasich 15, MOE ~5. So, the leaders are within the MOE.

Poll was 2 days ago, so the Fiorina thing was still hot, I think.

Then again:

Ted Cruz's Indiana plan: Throw everything at the wall and see what sticks - CNNPolitics.com

Cruz allies and people close to the campaign describe a budding sense of gloom, with internal polls diving as Trump mounted even stronger than expected showings in his native northeast. In Indiana, which Cruz backers once believed they were favored to win after his strong defeat of Trump in Wisconsin, Cruz's numbers have fallen precipitously: Once leading, Cruz now trails in the state by eight to 10 points, according to a person who has seen the numbers, with Trump over the 40% mark. Cruz's campaign did not respond when asked about those figures.

 

[T.Lex]

Yeah, but I take leaks of "internal polling" with a grain of salt (although less than you use when digesting one-on-one nominee polls).

There's a strange dynamic where leaking bad news helps set low expectations. Primaries are about exceeding expectations. One way to do that is to do really well, like Trump has done. The other is to set low expectations. Its kinda like what big companies do before they release earnings data - sometimes they release bad news to lower expectations.

Cruz, as the story notes, was leading in some polls in Indiana early on. He probably won't do as well as those early optimistic polls, so there is a reason for them to downplay good news.

The poll I linked appears to be neutral, and asked what I thought were good questions. Some of the other questions also appeared to validate other observations - such as how both candidates are both loved and hated. Not too much in the middle.

 

[BugI02]

They have formulas that factor that into account. It's part of the margin of error. There was a poll posted on here recently that put Trump up close to or possibly beating Hillary in winning the general. But it had over a 14 point MoE. In large part due to how the participants were selected. The actual polling was done by survey monkey which has a pool of those that sign up to take online surveys/polls.

Sigh. The mathematics that underly statistical sampling are based on using a random (in the meaningful statistical sense) sample. Selection Bias distorts the sample being used from ideal in ways that cannot be completely factored out or minimized. So when you say 'formula' if you are referring to rigorous mathematics then you are most likely addressing techniques to minimize MoE/Confidence Windows or correctly match desired MoE to sample size.

Perhaps this will help.

"Since some people do not answer calls from strangers, or refuse to answer the poll, poll samples may not be representative samples from a population due to a non-response bias. Response rates have been declining, and are down to about 10% in recent years. Because of this selection bias, the characteristics of those who agree to be interviewed may be markedly different from those who decline. That is, the actual sample is a biased version of the universe the pollster wants to analyze. In these cases, bias introduces new errors, one way or the other, that are in addition to errors caused by sample size. Error due to bias does not become smaller with larger sample sizes, because taking a larger sample size simply repeats the same mistake on a larger scale. If the people who refuse to answer, or are never reached, have the same characteristics as the people who do answer, then the final results should be unbiased. If the people who do not answer have different opinions then there is bias in the results. In terms of election polls, studies suggest that bias effects are small, but each polling firm has its own techniques for adjusting weights to minimize selection bias."

This was what I was referring to when I spoke of poll respondents being self selecting. It is difficult to adjust for selection bias because you don't know what you don't know about the pool of respondents. If you don't have a complete description of the necessary demographic make-up of an ideal sample and your field sample, you don't know what you're missing demographically nor how to adjust it to more closely approximate ideal. The 'techniques' of adjustment mentioned in the last line are not based in rigorous theory but are rules of thumb and ad-hoc tweaks, proprietary and unique to each polling firm and never peer reviewed

 

[BugI02]

Trump's America First has ugly echoes from U.S. history - CNN.com

Because nobody on the Trump campaign knows how to use Google? Yeah, that's it...that gotta be it..........doesn't it?

You know what, nobody really gives a ****. The run-up to WWII might as well have been a hundred years ago, ancient history. If America First best describes the policy one wants to pursue, why not take back the name from ancient history. Only the terminally PC will object. You know, the same people who are afraid to use the word crusade because it might upset the people plotting our demise

 

[BugI02]

You're not hearing American Nazi dog whistles are you? 'Cause I don't think Trump needs to pander to the White Supremacist vote, and that's the only sinister spin that could be put on that name selection.

 

[GodFearinGunTotin]

You know what, nobody really gives a ****.

Yes, I know.

 

[BugI02]

Sorry, GFGT. Way too harsh and you don't deserve it. Work had me all stirred up, single malt has been applied. I should have taken my 'meds' before INGOing up. Please accept my humble apology.

 

[ArcadiaGP]

New poll finds Cruz with double-digit lead in Indiana | TheHill

I've no idea what a suitable sample size should be with these sorts of polls... but 400 feels too small.

That would be .0001% of 2012's Indiana voter turnout.... which is likely to be blown out of the water in 2016.

 

[GodFearinGunTotin]

Sorry, GFGT. Way too harsh and you don't deserve it. Work had me all stirred up, single malt has been applied. I should have taken my 'meds' before INGOing up. Please accept my humble apology.

No sweat, brother.

 

[GodFearinGunTotin]

New poll finds Cruz with double-digit lead in Indiana | TheHill

I've no idea what a suitable sample size should be with these sorts of polls... but 400 feels too small.

That would be .0001% of 2012's Indiana voter turnout.... which is likely to be blown out of the water in 2016.

I do know my phone and cell phone has received calls from a number of unknown (to me) numbers in the last couple of weeks. I've wondered if most of those weren't likely to be pollsters.

 

[chipbennett]

New poll finds Cruz with double-digit lead in Indiana | TheHill

I've no idea what a suitable sample size should be with these sorts of polls... but 400 feels too small.

That would be .0001% of 2012's Indiana voter turnout.... which is likely to be blown out of the water in 2016.

...among registered voters...

Seriously? 4 days before the primary, and they're polling "registered voters"?

 

[chipbennett]

Also, there's this:

American Research Group
Indiana Republican Presidential Primary

Presidential primary preference:

Indiana
Likely Republican Primary Voters	Apr 27-28
Cruz	32%
Kasich	21%
Trump	41%
Other	*
Undecided	6%
Read this table down *= Less than 1/2 of 1% - = None