EPTUS Politics Mega-thread - Page 8139
| Forum Index > Closed |
Read the rules in the OP before posting, please. In order to ensure that this thread continues to meet TL standards and follows the proper guidelines, we will be enforcing the rules in the OP more strictly. Be sure to give them a re-read to refresh your memory! The vast majority of you are contributing in a healthy way, keep it up! NOTE: When providing a source, explain why you feel it is relevant and what purpose it adds to the discussion if it's not obvious. Also take note that unsubstantiated tweets/posts meant only to rekindle old arguments can result in a mod action. | ||
|
ticklishmusic
United States15977 Posts
| ||
|
ShoCkeyy
7815 Posts
| ||
|
Logo
United States7542 Posts
On July 20 2017 23:35 ZerOCoolSC2 wrote: How can you make slavery an institution, with 2 civil wars already done, and not have it lambasted? As a black male, I'd be hard pressed to even attempt to watch it. That just...hurts. I understand all of the political and current conditions of society playing a role, but it's a dumb move. There are other ways to make this without using slavery of blacks (assumption) into a staple of an HBO series. Edit2: Harry Harrison did it with his trilogy. The Red Rising trilogy also does a good job of doing what they are proposing. Maybe I'm being overly sensitive. I can't even make it to being offended at the dumb idea of slavery persisting. I'm baffled at how dumb "the south seceded" and "America is on the brink of its third civil war" sound together. If the south is its own country how can the future wars be civil wars? | ||
|
ZerOCoolSC2
8982 Posts
| ||
|
TheTenthDoc
United States9561 Posts
On July 20 2017 12:34 mozoku wrote: I don't doubt that you're correct (that there's a pay gap favoring males), but the bold part is actually not true. Academic journals don't publish negative results, and it's a point of serious criticism among researchers/scientists. If an academic does a study and find a result that doesn't find a significant effect (i.e. no significant pay gap), or an effect that confirms the status quo knowledge ("men make more than women"), then journals simply won't publish it. There's pretty big problems with the cycle this creates when it happens on a large scale: 1) Public gets interested in whether Assertion A on Topic X is true 2) Academics rush to study whether Assertion A on Topic X is true (due to how incentives are set up in academia). Let's say 1000 people study it. 3) Even if you assume researchers are playing it straight so you take p-values at their nominal level (a very bad assumption, btw), and you assume zero effect, you expect 50 of those 1000 studies (at the standard p < 0.05 p-value threshold that journals use) to produce positive results of Assertion A. 4) Journals only publish positive results (i.e. the 50 instead of the 950) of Assertion A. 5) The media reports that Assertion A is confirmed by academic research. If you look at the input and output of the system, what's happening is that public interest in a controversial claim literally produces research to support the controversial claim because of how the current academic process works. Of course, publications produced this way don't hold up during replication attempts. Hence why we have a replication crisis going on. Sorry for missing this earlier, but you're missing my point. This doesn't have to do with negative (i.e. nonsignificant) results. This has to with the properties of the null hypothesis that there is no pay gap and two-sided null hypotheses. Everyone in these fields is studying two-sided null hypotheses (because of poor reasoning and ease of doing it with regression models, but it's the truth). Even if only "significant" results are published, of those 1000 studies where 50 rejected the null spuriously half of them would be right-sided rejections and half would be left-sided rejections. If they were testing the hypothesis that women are paid less than men, it would be another thing entirely, but nobody is actually doing that-they're testing the null hypothesis that both genders get equal pay. This actually applies in all scenarios you describe where people are using two-sided hypothesis testing-you should find an equal number of "publishable" (i.e. significant) results on each side of the null value. You require an additional step 6 where journals are gagging results that run counter to the current evidence to really explain something like this. (this is part of why the "replication crisis" is really an effect exaggeration crisis, rather than an effect generation crisis; when you censor non-significant results for anything besides the null hypothesis, you bias the effect you estimate away from the null so subsequent studies are powered improperly, but a null effect is estimated without bias) | ||
|
ZerOCoolSC2
8982 Posts
President Trump did not do much to sell the Senate health care bill before its failure. But he gave the sale a shot Wednesday in the White House before cameras and a captive audience of nearly all the Republican senators. His comments were at times confusing, and in some cases, outright incorrect. It shows the challenge for a president who doesn't dive deeply into policy to sell his agenda. Here's a look at everything Trump said, with some fact checks and context: Source | ||
KwarK
United States42638 Posts
| ||
|
Plansix
United States60190 Posts
On July 20 2017 23:43 ShoCkeyy wrote: You guys obviously haven't seen The Man in the High Castle on Amazon... It's about what if the Axis powers did win the war. It was a hit show for it being on amazon. I'm sure if anybody can pull off an even better version, it would be HBO. You are right, but I think people are reacting to the creators and their track record. The man from high castle is a look at fascism as it would exist in the US. It is successful in its attempt and addresses the banal nightmare that is a fascist state. Look at how Germany addressed its post Nazi history and compare that to how the South addressed the post civil war era. I don’t think the GOT show runners have the chops to address the subject. | ||
|
ZerOCoolSC2
8982 Posts
On July 20 2017 23:56 Plansix wrote: You are right, but I think people are reacting to the creators and their track record. The man from high castle is a look at fascism as it would exist in the US. It is successful in its attempt and addresses the banal nightmare that is a fascist state. Look at how Germany addressed its post Nazi history and compare that to how the South addressed the post civil war era. I don’t think the GOT show runners have the chops to address the subject. The only reason I would even look at a trailer would be to see the stark contrast between the north and south. That would be all. If you read up on the books I mentioned previously, they tackle the exact same topic in a different manner and succeed imo. | ||
|
thePunGun
598 Posts
| ||
|
mozoku
United States708 Posts
On July 20 2017 22:45 Acrofales wrote: This actually isn't true at all. That isn't what people mean when they criticize journals for not publishing negative results. Negative results are if someone has an interesting hypothesis, so lets say "women get paid less than men". They do the research, and find that their hypothesis doesn't hold (or at least, the difference is statistically insignificant given the data available). They send it to a journal and the journal rejects the paper, because while the research is novel, the reviewers don't think the research is significant at all: clearly these researchers were dumb to pick this hypothesis in the first place! No offense, but that's the same thing as what I said: If an academic does a study and find a result that doesn't find a significant effect (i.e. no significant pay gap), or an effect that confirms the status quo knowledge ("men make more than women"), then journals simply won't publish it. Negative results are those without a significant effect, or those where the journal doesn't care about the significant effect. Nobody does a study hypothesizing "Men make more than women" for that reason. Years pass, and other researchers pick up on the topic and some find significant results in favour. This gets pulbished and generates a lot of buzz. The original researchers whose paper was rejected (or other researchers with similar data) now get in on it and say "wait a minute, I tested that, and there was no statistical significance one way or the other". They now make some changes in the manuscript that was rejected to show that they are invalidating prior research which did find an effect in their own data set, and send it for publication. It now gets published and generates quite a lot of buzz, because it is no longer a research paper that proposes a stupid hypothesis that turns out to be false, herpaderp. No, it is now a research paper that contradicts a popular hypothesis, which is valuable research. Eventually with enough studies, someone will gather all the data and do some meta-analysis, and average out all the different subgroup analyses for and against the hypothesis and say something one way or the other about the country as a whole (and other people will argue that that statistical analysis is wrong, and a different method should be used, which finds a different effect, etc. etc. etc.), but these are not the negative results that don't get published. These negative results definitely do get published. At that point, the "negative studies" are acting effectively as time-backwards failed replications though, which I did point out. The fact that academic journal publish a lot of research that fails attempted replication is no secret. While we're on the topic, you have a lot more faith in meta-studies as I do. It's well-known that researchers can pretty reliably produce "significant" results from effects with a true effect size of zero.Source (among many others) In a topic as politicized as the wage gap, it's almost certain that there's a lot of p-hacking, garden of forking paths, favorable data coding, sample selection, etc. going on to prod the data to say what the researcher wanted to find in the first place. It's well-known these things are abused whenever there are studies on other politicized topics. A meta-study is only as good as the studies it's aggregating. I'm not saying that meta-studies are useless or shouldn't be done, but they aren't a silver bullet. When done properly, they can be very valuable. The upshot is that you should never really trust a study (meta- or otherwise) unless until you read and understand how it was done. And even then, you should probably wait for replications (maybe not if it the study methodology was pre-registered). The journal publication threshold (including peer review) isn't a reliable indicator of quality on its own. Another factor (which is not really journals or academics' fault) is that the media tends to report novel findings ("Women found to get paid less than men"), but rarely report failed replications (backwards-time or not). Which leads to the public getting a distorted picture where controversial and counter-intuitive claims are reported to have a lot more evidence supporting them than there is in reality. This is a horrible mess and hurts society in a bunch of ways. Academia badly needs reforming. The problem with negative results not getting published is in the first part: the paper was rejected because it found no significant result because the reviewers couldn't argue that their hypothesis was interesting to study in the first place. But when is a hypothesis "uninteresting"? What if a (small) group of other researchers are actually also interested in this hypothesis. The same study keeps getting repeated (wasting resources), never finding significant results, and never getting published. For instance, someone might be studying the effect of peanut butter on the rotation of the earth. They find no result, and get laughed away from the journal. Two months later, someone else, unrelated, is also interested in whether peanut butter has any effect on the rotation of the earth... etc. Wouldn't it be better if at some point, someone would accept the paper showing that peanut butter has no effect on the rotation of the earth, and allow all these crazy peanut-butter-rotation-scientists to get on and study something useful? Agreed here, and that's another common criticism of not publishing negative results. Are you an academic? I could see why academics (presumably the community you spend a lot of time in) complain more about your last paragraph (it directly affects them), and less about the issue in my posts (it doesn't really affect them, and talking about it could potentially invalidate their work). Overall, I don't think were very far apart in what we said. So I'm a little confused on how you started with "This actually isn't true at all. That isn't what people mean when they criticize journals for not publishing negative results." | ||
|
ZerOCoolSC2
8982 Posts
On July 20 2017 23:59 thePunGun wrote: Uhm, it's not like Lincoln was an abolitionist or even believed blacks should be granted the same rights as whites. One of Lincoln's key goals during the Civil War was to keep England from supporting the South, which is another reasons, why he thought emancipation was the right move. Because the British crown changed their stance on slavery. (Source) We're not talking about what Lincoln believed in. We're talking about how dumb a show this would be. | ||
|
KwarK
United States42638 Posts
On July 20 2017 23:59 thePunGun wrote: Uhm, it's not like Lincoln was an abolitionist or even believed blacks should be granted the same rights as whites. One of Lincoln's key goals during the Civil War was to keep England from supporting the South, which is another reasons, why he thought emancipation was the right move. Because the British crown changed their stance on slavery. (Source) Britain had zero interest in supporting the south. | ||
|
{CC}StealthBlue
United States41117 Posts
| ||
|
Doodsmack
United States7224 Posts
| ||
|
TheLordofAwesome
Korea (South)2617 Posts
On July 21 2017 00:14 Doodsmack wrote: Dat timing https://twitter.com/aaronblake/status/888044710408970240 I was just about to post this. | ||
|
{CC}StealthBlue
United States41117 Posts
| ||
|
On_Slaught
United States12190 Posts
| ||
|
TheTenthDoc
United States9561 Posts
On July 21 2017 00:14 Doodsmack wrote: Dat timing https://twitter.com/aaronblake/status/888044710408970240 "Come at me bro" seems an apt description (though this scores pretty low on the "anonymous source reliability" scale 538 talked about). | ||
|
thePunGun
598 Posts
On July 21 2017 00:08 ZerOCoolSC2 wrote: We're not talking about what Lincoln believed in. We're talking about how dumb a show this would be. Well, isn't it premature to judge a show, before the pilot even aired. I'll at least give it a fighting chance. If the writing's good, the show might be aswell. ----- edit: On July 21 2017 00:09 KwarK wrote: Britain had zero interest in supporting the south. That's not entirely true, most of the british elite supported the confederacy, but Lincoln's Emancipation Proclamation according to wikipedia"caused European intervention on the side of the South to be unpopular." ----- Since this is the politics thread after all...Has anyone read the new CNN headline: "991 tweets, 0 pieces of major legislation Trump's first six months in office by the numbers" Angry tweet no. 992 incoming...  | ||
| ||
