The Classical Freedom loving Left vs the Regressive Leftists

[Princeps]

Perhaps you should pull your head out of the sand and start addressing the issue as the massive civilisational problem it is

Sure, but it has nothing to do with theology or religion itself.

Let will repeat myself:

Most muslim countries are poor, rural and mired by a political culture of corruption and despotism. This leads to a tendency called preference falsification:

https://en.wikipedia.org/wiki/Preference_falsification

There is a lack of public knowledge of the advantages or the meaning of liberalism and secularism in these countries. The corrupt agency culture also makes improvements slow and difficult (i.e. it's difficult to convince people of the benefit of liberal public goods when the law enforcement expected to provide them is only out for its own interests).

None of this is the result of Islamic theology, but the outcome of complicated historical events that have shaped the middle east.

It doesn't matter what the holy books say. Christianity has all the same ugliness in its underlying DNA: the difference is that European society developed differently from 1500s onwards, mainly thanks to geographic differences.

 

[Ryika]

Together, the surveys involved more than 38,000 face-to-face interviews in 80-plus languages and dialects, covering every country that has more than 10 million Muslims except for a handful (including China, India, Saudi Arabia and Syria) where political sensitivities or security concerns prevented opinion research among Muslims.

http://www.pewforum.org/files/2013/04/worlds-muslims-religion-politics-society-full-report.pdf

Seems to be a rather normal sample-size to me. Only problem I see is the "more than 10 million Muslims"-number, obviously that excluded a lot of the western world and therefor the more moderate Muslims.

But other than that I don't see any valid ground for "Somebody polled 1.62 billion people?"-nonsense. Obviously people didn't and you very much know how that works, Flying Pig - by polling a sufficiently large sample size, then making a projection for the overall population.

Still, as all graphic only shows the information that it wants to show and leaves out bits like:

Among sharia supporters, median % of Muslims
who say sharia should apply to Muslims only

Southern-Eastern Europe 64%
South Asia 60%
Central Asia 59%
Southeast Asia 55%
Middle East-North Africa 51%

The trends shown in the graphic however can definitely be found in the study itself.

 

[Flying Pig]

Extrapolating 38,000 to 1 billion means that every person gets to speak for 26,000 people. Taking a look at p. 150, they asked about 1,000 people from each country, which is pretty rubbish if you're trying to forecast an election, let alone characterise a religion. Also, not going to Saudi Arabia, with 30 million, or India, with 172 million, is pretty inexcusable.

 

[Ryika]

Extrapolating 38,000 to 1 billion means that every person gets to speak for 26,000 people.

Which is not a problem at all. As long as your survey does not have sampling errors and the sample size is large enough to give confident results scaling the results to a bigger population has little Effect on the margin of error. A sample size of 1000 per country is certainly not a number that gives a high confidence level for the individual country, but a sample size of 36000 overall is more than enough to give a stable statistic.

But even if we were to accept that the number of people polled is not representative of the overall population, what would that really mean? That we can't be sure about the exact numbers, however there is no way the trends indicated by the survey are incorrect.

 

[Funky]

Extrapolating 38,000 to 1 billion means that every person gets to speak for 26,000 people. Taking a look at p. 150, they asked about 1,000 people from each country, which is pretty rubbish if you're trying to forecast an election, let alone characterise a religion. Also, not going to Saudi Arabia, with 30 million, or India, with 172 million, is pretty inexcusable.

And here comes the usual regressive excuse tactic: Don't like the news? Denigrate the messenger! Of course polling data is not 100% accurate, and nobody claimed it is. But it will most likely be pretty darn close. Not only is Pew one of the most respected research centers in the world, the data it has acquired on Muslim attitudes is extensive. It also correlates with the data of every other survey done among Muslim populations. As for Saudi Arabia, the most extreme countries didn't even allow the polling to be done, so if anything expect the real numbers to be even higher than in the graph.

 

[Flying Pig]

Which is not a problem at all. As long as your survey does not have sampling errors and the sample size is large enough to give confident results scaling the results to a bigger population has little Effect on the margin of error. A sample size of 1000 per country is certainly not a number that gives a high confidence level for the individual country, but a sample size of 36000 overall is more than enough to give a stable statistic.

I disagree. If each individual sample is not reliable, they become less reliable, not more reliable, by aggregating them. If you run polls in Birmingham, London and Cardiff, and neither gives you evidence in which you're confident, you're a fool to be confident that you've got conclusions which can be applied to the whole country.

 

[BvBPL]

Which is why polls contain internal means to determine confidence. The error value of a poll is quite possible to determine based on the poll itself assuming the poll is well crafted.

 

[Flying Pig]

My contention is that the poll is not well crafted. Bangladesh has a Muslim population of about 150 million, and they asked less than 2000 of them, which would be recognised as a rubbish sample size (by YouGov, who won't go below 2000 to predict the attitudes of the 45 million voters in the UK). Then there's Indonesia, with over 200 million Muslims, of whom they asked about 2000. For Albania's <2 million Muslims, on the other hand, they asked just over 700. Doing the maths, that makes each Albanian vote worth over three times as much as each Indonesian vote. They also don't even try to touch huge and significant populations, such as the 170 million Muslims in India - who by themselves cancel out Albania, Azerbaijan, Bosnia, Kosovo, Lebanon, Russia (itself grossly over-sampled - each Russian has the same weight as over 17 Indonesians), Palestine, Iraq, and probably a fair few others. If the data collection is so bad, you can't set anything in store by the conclusions it draws. This isn't the same as saying that the conclusions it draws are necessarily false, only that it's totally untrustworthy.

 

[Ryika]

I disagree. If each individual sample is not reliable, they become less reliable, not more reliable, by aggregating them. If you run polls in Birmingham, London and Cardiff, and neither gives you evidence in which you're confident, you're a fool to be confident that you've got conclusions which can be applied to the whole country.

But each individual sample is not "not reliable". With a margin of error of 5% and a confidence level of 99% (most polls are fine with 95%) you can basically poll 1000 people and project the result on a near infinite amount of people.

A higher sample size is only needed if you're not okay with a margin of error of 5% (or want a higher confidence level).
But yes, you're right, adding the numbers would indeed decrease the confidency, not sure what I was thinking there.


Overall, what you're basically arguing against is this small percentage of variance, that is not at all unusual for polls. There is no way the trend shown in these polls can be incorrect, period. So why are you spending so much energy trying to find ways to deny its findings?

/edit - Speaking of that:

My contention is that the poll is not well crafted. Bangladesh has a Muslim population of about 150 million, and they asked less than 2000 of them, which would be recognised as a rubbish sample size (by YouGov, who won't go below 2000 to predict the attitudes of the 45 million voters in the UK). Then there's Indonesia, with over 200 million Muslims, of whom they asked about 2000. For Albania's <2 million Muslims, on the other hand, they asked just over 700. Doing the maths, that makes each Albanian vote worth over three times as much as each Indonesian vote. They also don't even try to touch huge and significant populations, such as the 170 million Muslims in India - who by themselves cancel out Albania, Azerbaijan, Bosnia, Kosovo, Lebanon, Russia (itself grossly over-sampled - each Russian has the same weight as over 17 Indonesians), Palestine, Iraq, and probably a fair few others. If the data collection is so bad, you can't set anything in store by the conclusions it draws. This isn't the same as saying that the conclusions it draws are necessarily false, only that it's totally untrustworthy.

See:

Q. How big are your samples &#8211; and how do these compare with those of polls using traditional methods?
A. Most of YouGov's polls of the general public achieve samples of at least 2,000. Among other companies conducting regular polls for the media, MORI&#8217;s monthly political tracking polls also have samples of around 2,000, while ICM and Populus normally poll 1,000 people. The headline voting intention figures - which eliminate respondents who say "don&#8217;t know" or "won't vote" or, in the case of MORI, ICM and Populus, are judged unlikely to vote &#8211; are based on samples, typically, of 1,500 for YouGov, 1,000 for MORI and around 550 for ICM and Populus.

Q. Why does sample size matter?
A. Because the risk of random sampling error is related to sample size: the smaller the sample, the greater the risk of such error. On a sample of 550, we can be sure that, 19 times out of 20, the true figure &#8211; that is, the figure that would have been obtained had the whole population been polled using the same methods &#8211; is within 4% of the published figure. Random error on a sample of 1,000 is up to 3%, on 1,500 up to 2.5% and on 2,000 up to 2%. Larger samples also allow the views of subgroups, such as women voters or Conservative supporters, to be measured more accurately.

https://yougov.co.uk/about/panel-methodology/research-qs/

 

[Flying Pig]

OK, but that doesn't account for the weighting between countries. Unless Muslims everywhere are some kind of hive-mind, we'd expect differences between how Albanian Muslims (recently the victims of genocide), Russian Muslims (only recognised officially since the 1990s) and Indonesian Muslims (the world's largest Muslim country) see the world. Why does the study weight the Albanians and Russians so much more heavily? Why does it draw conclusions about all of the world's Muslims based largely on countries with small Muslim populations, ignoring those with more population that most of the sampled countries put together? I refuse to believe the justification that India is too dangerous in which to poll, when they went to Afghanistan.

Also, note what YouGov say about subgroups - the logic only holds if the 1,000 people are taken from a population that is internally all the same. For example, if group A makes up 0.01% of the population, only 1 will be polled in a sample of 1000. The poll will then make his voice 'count' for (in the UK) 70,000 people.

 

[Funky]

Where is your problem? Every country is listed separately in the study. We can go into detail and examine regional differences. We can note that in Afghanistan 99% of the population are in favour of sharia, whereas in Indonesia it is "only" 72%. Or that in Indonesia 93% say the wife must obey her husband, whereas it is 65% in Turkey. The whole point of the graph is to sum up the findings and extrapolate to the entire Muslim population. Yes, there is some inaccuracy in this process. We can't tell if the real numbers are slightly lower or slightly higher. But this is the best data we have. It correlates with every other study done among Muslim populations. It is accurate enough to draw well-founded conclusions.

Yet you seem desperate to denigrate the study. Why? Because you don't like the findings? Because you don't want to accept that they present a huge problem? Wake up to reality, man. It's about time.

 

[Ryika]

OK, but that doesn't account for the weighting between countries. Unless Muslims everywhere are some kind of hive-mind, we'd expect differences between how Albanian Muslims (recently the victims of genocide), Russian Muslims (only recognised officially since the 1990s) and Indonesian Muslims (the world's largest Muslim country) see the world. Why does the study weight the Albanians and Russians so much more heavily? Why does it draw conclusions about all of the world's Muslims based largely on countries with small Muslim populations, ignoring those with more population that most of the sampled countries put together? I refuse to believe the justification that India is too dangerous in which to poll, when they went to Afghanistan.

I actually don't think the study does? I haven't read the whole thing, but the bits that I've read (and the graphs that I've looked at ) clearly separate by countries, as do all the tables at the end. There are some conclusions that are drawn about the "overall Muslim population", but those follow separately, after the numbers per country have already been named.

The graphic linked a few posts ago obviously does, but it's clearly indicated that it does exactly that. I don't see anything wrong with drawing an average.

Also, note what YouGov say about subgroups - the logic only holds if the 1,000 people are taken from a population that is internally all the same. For example, if group A makes up 0.01% of the population, only 1 will be polled in a sample of 1000. The poll will then make his voice 'count' for (in the UK) 70,000 people.

...which is once again perfectly fine and already factored into the numbers above. Yes, it does give some inaccuracy, but that's fine. Pretty much all polls are bound to be off-target by a few percent, we usually accept ~5%.

So this would only be relevant if either you want to get numbers that are more accurate than "a few percent off" (which gives heavy diminishing returns if you try to get <3% while keeping a high confidency), if you assume the sample was not properly randomized (which you would need evidence for, just assuming an error because you don't like the results - especially on this scale - would be a fallacy) or if a population is so diverse that 1000 people cannot truly represent all opinions (which is practically only possible with questions that have more than a few dozens possible answers or a blank field that you can fill in with whatever you want, but not with yes-or-no answers as used in the survey).

 

[Princeps]

Wake up to reality, man. It's about time.

There is nothing to wake up to.

Take ISIS. Most Muslim leaderships support Western efforts to destroy the organization. Most polling indicates majorities in many Muslim countries approve of the American campaign against ISIS (and even if they don't, the hatred of Daesh is deep all over the middle east).

In other words, even with their highly conservative views (borne from social pressure and choice blindness), most Muslims absolutely resent the Sharia of ISIS and their ilk and do not see them as a viable alternative to modernity.

http://www.pewglobal.org/2015/06/23...e-critical-of-post-911-torture/bop-report-28/

 

[illram]

The numbers are accurate, and can be looked up in the studies of every major research center. We have been posting the results of these studies for years. Perhaps you should pull your head out of the sand and start addressing the issue as the massive civilisational problem it is. Here is an overview of just a few problematic beliefs that Muslims hold, based on Pew research.

A while back you posted some bogus study about Muslims, and I pointed you to the Pew numbers you are now touting since they are more accurate, your misleading graphic notwithstanding. Here's that post for posterity. So you're welcome I guess?

We'll have to agree to disagree about my head being in the sand. I mean Funky, really, I am offended! I gave you the delicious cornucopia of Muslim polling data from which you have cherry picked ever since. I would expect some more recognition than that.

I recognize the issues in current popular Islamic orthodoxy, and I have done so before it became the soup de jour of the Scared White People community. I'm not a cultural relativist. Saudi Arabia is a terrible place. Iran's Ayatollah is a serious jerk, and is he trying to look like an evil Super villain? C'mon man smile for the camera at least.

I just don't see all of this through the us vs. them mindset you seem to have. For me "us" includes Muslims. If they have a particularly conservative mindset, as many people of other religions do that I disagree with then I will rarely see eye to eye with them on a variety of social and political issues and that's that. I don't see it as much more than that. I can segregate the very very small sliver of dangerous people, i.e. terrorists, ISIS footsoldiers, etc., from the larger mass of normal everyday Muslims regardless of how "out there" their political or religious beliefs are. I am not making the extra leap you are that lumps all the terrorists into the giant bag of all Muslims, and then declaring that since they're all in the same giant bag we need to deal with them all the same way.

 

[Funky]

A while back you posted some bogus study about Muslims, and I pointed you to the Pew numbers you are now touting since they are more accurate, your misleading graphic notwithstanding.

Please enlighten me how anything I have posted is "bogus" or "misleading". I am serious. By the way, the Pew results you posted don't contradict what I said, namely that vast numbers of Muslims support terrorism, in fact they confirm it. Oh, and what do you make of the over 80% of Arabs who support IS, according to an Al-Jazeera poll?

I gave you the delicious cornucopia of Muslim polling data from which you have cherry picked ever since. I would expect some more recognition than that.

No need to cherry pick. Pretty much all polling data comes to the same conclusions.

I just don't see all of this through the us vs. them mindset you seem to have.
For me "us" includes Muslims.

As long as these Muslims behave like secular, moral people it absolutely does for me too. I think this is how most people think. Do you know who doesn't tend to think like that? Who has a very strong sense of "us vs them"? Muslims. Not all. Not Maajid Nawaz. Not Hamed Abdel Samad. Not my college at school who is a good friend of mine. But huge portions of the Muslim world, who divide the world into friends and enemies, into Muslims and kafirs, do.

I can segregate the very very small sliver of dangerous people, i.e. terrorists, ISIS footsoldiers, etc., from the larger mass of normal everyday Muslims regardless of how "out there" their political or religious beliefs are. I am not making the extra leap you are that lumps all the terrorists into the giant bag of all Muslims, and then declaring that since they're all in the same giant bag we need to deal with them all the same way.

Terrorism and jihadism is only the tip of the iceberg. Outside of jihadists we have around 20% of Muslims who are Islamists, people whose life goal it is to spread Islam around the globe. And outside of this group, most of your "normal, everyday Muslims" think that wives must obey their husbands and that sharia should rule. That doesn't mean they should be treated the same as terrorists (obviously). It does mean that we have a huge problem with our Muslim population, as can be observed throughout the Islamic world and in the Western Muslim diaspora every day.

As I tell everyone, do your research. Don't take anything I post for granted. I don't want to be mistaken about anything longer than I have to be. If you have data that proves anything I say wrong, then please provide it. Just stop falling back to ostrich behaviour and snarky comments, as you have done previously, since that helps nobody.

 

[Ajidica]

I have to say; I love how when you mentioned the Al-Jazeera poll you linked not to Al-Jazeera but instead to the Express - which helpfully neglected to include much in the way of linking back to the Al-Jazeera poll itself.

Plus, I think the Express's math may be off:

More than eight out of 10 of the respondents to an online survey for the Qatar-based network &#8211; which also has offices in London &#8211; gave their backing to the brutal regime.

The poll, which asked in Arabic &#8220;Do you support the organising victories of the Islamic State in Iraq and Syria (ISIS)?&#8221; has attracted almost 60,000 votes.

Shockingly, 81.3 per cent &#8211; or 61,145 people &#8211; voted yes to the question, in support of the Islamic extremists.

To get 81% of the poll respondents supporting IS and have the number of people voting in favor be 61145 they would have needed roughly 75200 respondents in total - not particularly close to "almost 60000 people". Plus there is the whole issue with online polls especially for less developed regions of the world.
Lastly, I wonder what "organizing victories" actually means. Is the poll referring to IS establishing a bureaucracy and civil administration?

 

[Funky]

I have to say; I love how when you mentioned the Al-Jazeera poll you linked not to Al-Jazeera but instead to the Express - which helpfully neglected to include much in the way of linking back to the Al-Jazeera poll itself.

Who cares? The data is the same. The figure has been reported across the media spectrum.

Lastly, I wonder what "organizing victories" actually means. Is the poll referring to IS establishing a bureaucracy and civil administration?

Does it make a difference? Do you set the bar so low? Should we not expect every sane moral person to be wholeheartedly against IS?

 

[sophie]

Wait, his evidence is based on a public internet poll?

AHAHAHAHHAA

 

[Aea]

Wait, his evidence is based on a public internet poll?



AHAHAHAHHAA

These are perfectly valid sources. When they agree with your bias.


Sent from my iPhone using Tapatalk

 

[Ryika]

These are perfectly valid sources. When they agree with your bias.

Unfortunately that's true in way too many cases.

Really is of course that basically any poll that is open for people to opt-in is practically useless because of (self-)selection bias. A person who is willing to trust such polls when they agree with them and disregards those that disagree with them should check their priorities, because that's a clear signal of a mind that is lying to itself to be able to maintain their view on the world.

I know that probably didn't need to be said, but... hey, I said it anyways.