President Donald Trump
I assume it's still acceptable to have a Trump thread in a Politics forum?
So this is an obvious lie - basically aimed at
no you didn't. you on purpose insisted about something I didn't write, attributed it to me, and kept going for your Marxist tangent on that regardless of facts, like all leftists do.
Nope. Here is the quote
btw the "non random sampling" has to be treated as random unless you can define the structure (skew) of the non randomness.
That’s just wrong and betrays a profound ignorance of how random sampling works. Call it Marxist if you want but it’s pretty basic math.
That’s just wrong and betrays a profound ignorance of how random sampling works. Call it Marxist if you want but it’s pretty basic math.
it isn't (and it isn't basic statistics).
if you don't know anything about a sample other than "it is non random", you still have to treat it as random, while acknowledging the information value of the sample will be lower than a proper random sample.
you can also entirely discard it if you can afford to do so: but often enough you can't sample randomly in a proper way to avoid the problem.
so you think, suggest, even "know" the sample is non random, you still have to treat it as random unless you know how and why it is non random.
it isn't about a lack of knowledge on how to randomly sample when you can do that.
it's about what to do with imperfect data sets.
in economics it's extremely common to treat samples as random even if you intuit they might not be so for various reasons, only do add all those caveats in the discussion.
because it's not like you can go and resample randomly what happened at a certain point in time in a country, do you understand that?
you come from a field where you can resample. fine when you can, ofc you will do that and throw away non random samples.
when the only data sets you have to try to learn SOMETHING about reality are potentially non random samples (because you don't know how response rates distribute or anything else), I suppose in your field you just give up. in economics we don't
ok so which data would you use to claim that college graduated people voted democrat more than republican in 2010?
Mitt’s not the smartest guy in the world but I’m sure he was not dumb enough to expect 2010 turnout demos to be a good proxy for 2012.
Mitt’s not the smartest guy in the world but I’m sure he was not dumb enough to expect 2010 turnout demos to be a good proxy for 2012.
college educated people tend to vote more often than the rest of the population.
which means that if you know how they skew in 2010, that's still very predictive of how they will skew in 2012 (unless ofc candidates and conditions are different), because most of them vote in non presidential elections as well.
the rest of the population does have big changes in turnout during presidential elections. but that is not the people we wanted to assess right?
so the 2010 skew in college educated voters would still have been the best predictor of the skew in college educated voters in 2012. better an 2008.
it's just that a meaningful amount of college educated voters do like to pay for parasites. that's your choice.
and Romney was politically stupid not to know that, sure.
it remains factually true that when he said the 47% thing, that was factually right. except he forgot people (so far away from his morals that he could have a model of the mind for them) exist that like the idea of paying taxes for parasites
it isn't (and it isn't basic statistics).
if you don't know anything about a sample other than "it is non random", you still have to treat it as random, while acknowledging the information value of the sample will be lower than a proper random sample.
Nope. It’s actually the exact opposite. The vast majority of selected small samples from a large population will not converge to population means. Random samples are one of the very few that do and that’s one of the main reasons they are used. Trying to extend a very very special property of random samples to others is fallacious. Again, people with no math training even will agree with me on this at it is pretty basic.
Nope. It’s actually the exact opposite. The vast majority of selected small samples from a large population will not converge to population means. Random samples are one of the very few that do and that’s one of the main reasons they are used. Trying to extend a very very special property of random samples to others is fallacious. Again, people with no math training even w
ofc they won't and that wasn't my claim.
my claim is just that an added layer of uncertainty to a sample doesn't make it completely worthless and you can still use it pretending it is random (when treating it) then adding the caveat you have reasons to believe it wasn't random so the confidence interval basically should be "wider".
you are failing to understand the context.
the context is you want to know something about a presumably stochastic dynamic but you only have samples that could be inherently biased in various ways (you aren't sure which ways though).
you can either not touchy the subject , or touch it acknowledging that and treating the sample as random, but with lower information value because actually they might not be random (but you don't know how they skew).
again it's a field issue.
you are in physics if I understand it correctly, the only case the topic might apply to you is: you are given top secret data gathered through espionage of some sampling from a lab in China but you have doubts about the data gathering methodology.
is there still any information you can extract from that? because in normal instances you just ****ing runnthe experiment again, or call the people who conducted it.
now on economics, particularly in macro, you can't rerun ****.
you have these partial samples of actuarial data from inheritances in Frankfurt in the 17th century and you want to know how often the inheritance included something more than the main house.
you know, or you suspect, recorded inheritances aren't a random sample of inheritances because of local laws at the time discriminating (say) against women heirs or Jews. you know or suspect different subdemographics wouldn't report all in their inheritances (because say taxes).
now can you still extract any useful information from that partial, biased, datasets of Frankfurt inheritances? yes/no? if yes how do you treat it? as random!
and then you add all the caveats admitting your economic history finding about German inheritances are only very partial because xyz
when he first showed up i liked his schtick - then when it was obvious how partisan he was i feel like that took away any credibility he had - ie he wasn't trying to expose bad government, just expose the liberal side of it
so i like where this is going - i've long been rooting for the fracturing of the 2 party system and always held out some hope that a silver lining of trump 2.0 could be just that
I did explain it but you claimed not to understand. I can tell it to you, I can’t understand it for you…
I asked you to define a term and you didn’t want to. I can’t make sense out of a statement where one of the terms is unclear to me. We might not even be disagreeing if I don’t know what you mean.
And anyway, it wasn’t this exchange that made me decide. It was back when you were losing your **** over men’s rights and trans people.
I think it was men’s issues we were talking about? It could have been men’s rights but I don’t particularly think men’s problems are a lack of rights.
But yeah, I do think there’s a certain amount of uncomfortable associations between radfems and anti-trans sentiments. I don’t think I made any particular accusation about you personally being anti-trans.
What you wrote above is certainly true to a degree. But what you wrote below isn't the only cause.In some cases, liberals have perpetuated the stereotype by writing things or saying things that sound relatively elitist, or that at a minimum are easy to mischaracterize as expressions of elitism. The endless discussion of What's the Matter with Kansas and Obama's gaffe about wo
Completely agree.
It’s more of the broader point that Dems in general are tiptoeing around this coastal elite stereotype and it’s partly due to discomfort with the nuggets of truth in the stereotype.
But it’s not only due to that. As it is not only due to lies that Republicans perpetuate.
This is a a juvenile way of thinking about politics and the media imo. You aren't "getting to" Fox just because Fox is reporting negatively on what you are doing. That's just what the outrage vendors on the left side of the spectrum want you to believe.
Steamraise likes posting that kind of stuff for some reason.
Anyone expecting a good-faith, rational discussion with a hater is probably going to be disappointed.
(Just making a general observation; it may or may not be applicable to any actual persons or conversations currently taking place in this Forum.)
That’s not bait. If I want to bait you, I’ll fully chum the water, don’t worry. It’ll be obvious even to the likes of you, cutiepie 😀
I just don’t know what you’re talking about weirdo
ofc they won't and that wasn't my claim.
my claim is just that an added layer of uncertainty to a sample doesn't make it completely worthless and you can still use it pretending it is random (when treating it) then adding the caveat you have reasons to believe it wasn't random so the confidence interval basically should be "wider".
I quoted your claim exactly. It was wrong.
Breaking news (sorry if already shared):
Donald Trump, who claims to be the most anti-war president ever and to have already stopped 7 wars, is renaming the Defense Department to the Department Of War.
Oh, and yesterday he bombed some ships. Seems to be spoiling for a fight.
Nope it isn't. My solution for the fertility problem is to accept it and stay out of it because nothing constitutional can be done for it.My solution for people not reading enough books is to accept itMy solution for trolls like you in social media is to accept they existMy solution for people putting pineapple on pizza is to accept itMy solution for AI being potentially lethal
yes
The only solution u don’t accept is left wing people and immigrants existing and should be put to death …
I certainly am not in favor of implicitly endorsing GOP narratives about coastal elites. But I also don't want to ignore the political sentiment in the country to the detriment of winning elections. I don't think you have to nominate Gavin Newsom in order to take on bullshit narratives about coastal elites.
Democrats often walk on eggshells, afraid that any mistake will cost them the election. The problem is that while there are genuine political calculations to be made, this continued insistence to protect everyone's sensibilities just comes off as timid, afraid and annoyingly apologetic.
Democrats often walk on eggshells, afraid that any mistake will cost them the election. The problem is that while there are genuine political calculations to be made, this continued insistence to protect everyone's sensibilities just comes off as timid, afraid and annoyingly apologetic.
What about my post had anything to do with protecting people's sensibilities?
I was commenting more broadly on the issue of coastal elites. For the discussion to make sense, a fair bit of voters out there must be deathly afraid, angry or annoyed by coastal elites. Or rather by the people they are told are coastal elites.
However, the current crop of GOP politicians sway voters with tales of dog-eating immigrants, anti-vax conspiracies, election fraud lies and supposed UFO coverups. The messaging doesn't have to be rational, emotional is good enough.
You could roll out a hillbilly from the most dialect-ridden parts of Appalachia, and the campaign response would merely be that he sold out to the coastal elites.
ofc they won't and that wasn't my claim.
my claim is just that an added layer of uncertainty to a sample doesn't make it completely worthless and you can still use it pretending it is random (when treating it) then adding the caveat you have reasons to believe it wasn't random so the confidence interval basically should be "wider".
I'll just close out the basic math in case anybody is still wondering. Nobody said that. Again here is what you wrote.
btw the "non random sampling" has to be treated as random unless you can define the structure (skew) of the non randomness.
This is wrong because you rarely if ever know how skewed nonrandom data (or even random data is for that matter) because you don't actually know what the population level data is. That's the whole reason you're having to use a small subset population in the first place. In reality there are a whole host of techniques on how to draw conclusions from non randomly generated data and it's not simply to assume random with a larger interval lol. I stand by my claim that this is basic high school level stuff, outside of right wing culture.
