IQ and religion

[Sizwe]

Judging by the comments, the No faith - Higher IQ correlation clearly does not apply here..

[OB1]

Judging by the comments, the No faith - Higher IQ correlation clearly does not apply here..

In reality it applies nowhere.
The more people appear in your statistic the more it will even out.
In the end it's going to be even.
Unless you manipulated it somehow which I think this particular statistic is.

Also IQ doesn't tell that much.

[NordicNerd]

In reality it applies nowhere.
The more people appear in your statistic the more it will even out.
In the end it's going to be even.
Unless you manipulated it somehow which I think this particular statistic is.

Also IQ doesn't tell that much.

Wrong on all accounts! A statistical relation in the form of a correlation, regression or some other association measure does not "even" out with increasing sample size. Two things happen in such a case:

1. The estimate of the strength and direction of the association becomes more precise (the statistical model becomes more accurate)
2. The generalizability of the association from sample to population becomes more robust (lower p values).

Thus, increasing sample size leads to the opposite of what you suggest.

And- IQ is one of the most important and robust predictors of academic and work performance as well as socio-economic status.

NN

[OB1]

Wrong on all accounts! A statistical relation in the form of a correlation, regression or some other association measure does not "even" out with increasing sample size. Two things happen in such a case:

1. The estimate of the strength and direction of the association becomes more precise (the statistical model becomes more precise)
2. The generalizability of the association from sample to population becomes more robust (lower p values).


And- IQ is one of the most important and robust predictors of academic and work performance as well as socio-economic status.

NN

^^Even out in terms of accuracy.

Example:
I have 3 people in my statistic.
1 is smart and atheist.
2 are not that smart and believe in religion.
This will not reflect the actual state of affairs.
Therefore you should have maybe 100 of each nation at least.

Regarding IQ:

IQ scores not accurate marker of intelligence, study shows

[Alfurinn]

I don't think religious people are less intelligent. Some weeks ago I read an article that says religious people are more intelligent than non-believers...  

Other more interesting finding is that there's a region in the brain where critical thinking takes place, I think it is called the hippocampus... it shows less activity in religious (fundamentalist) people compared to that of non-believers. Apparently, because of this, their hippocampus shrinks a bit.

That's why people with rigid religious beliefs have problems to question their beliefs or accept it when others question their beliefs.

[el numero uno]

I fail to see what P-values have to do with correlation, this is not a test of significance.

[Griffith]

The dumb shall inherit the Earth.

[NordicNerd]

^^Even out in terms of accuracy.

Example:
I have 3 people in my statistic.
1 is smart and atheist.
2 are not that smart and believe in religion.
This will not reflect the actual state of affairs.
Therefore you should have maybe 100 of each nation at least.

Regarding IQ:

IQ scores not accurate marker of intelligence, study shows

The meta-analytic approach deals with these matters- in fact that is the point of it. : https://en.wikipedia.org/wiki/Meta-analysis

NN

[NordicNerd]

I fail to see what P-values have to do with correlation, this is not a test of significance.

As I wrote: increasing sample size leads to less bias and more precision regarding the estimate of the correlation. Further, it leads to less uncertainty regarding generalizability from sample to population through diminishing p-values. (less chance of both type 1 and type 2 errors).

NN

[el numero uno]

As I wrote: increasing sample size leads to less bias and more precision regarding the estimate of the correlation. Further, it leads to less uncertainty regarding generalizability from sample to population through diminishing p-values. (less chance of both type 1 and type 2 errors).

NN

Nope. A bigger sample means the data will be closer to the normal curve (Central limit theorem) thus when you use it to calculate percentages, probabilities, etc, results will be more accurate. This doesn't translate in lower p-values, for instance, you can have a n = 1000 perfectly fitting the normal curve, but if there's not statistical significance between the things being measured, then P-values aren't going to be on the low side.

Increasing the sample size doesn't result in lower p-values.

[NordicNerd]

Nope. A bigger sample means the data will be closer to the normal curve (Central limit theorem) thus when you use it to calculate percentages, probabilities, etc, results will be more accurate. This doesn't translate in lower p-values, for instance, you can have a n = 1000 perfectly fitting the normal curve, but if there's not statistical significance between the things being measured, then P-values aren't going to be on the low side.

Increasing the sample size doesn't result in lower p-values.

You are wrong. The parameter estimates of the model becomes more accurate and the p-values gets smaller with increasing sample size, assuming there is a true population effect. This is basic inferential statistics. This is why p-values are of little interest in very large samples- practically irrelevantly small effects (associations, differences of means etc) become significant.

Further, although not very important for my argument, what you say about the normal curve is also wrong as not all natural distributions follow a gaussian distribution. For instance event-data, such as number of road accidents in a given time-period, are Poisson-distributed.

NN

[el numero uno]

You are wrong. The parameter estimates of the model becomes more accurate and the p-values gets smaller with increasing sample size, assuming there is a true population effect. This is basic inferential statistics.

Further, what you say about the normal curve is also wrong as not all natural distributions follow a gaussian distribution. For instance event-data, such as number of road accidents in a given time-period, are Poisson-distributed.

NN

Sure, not every set of data follows the normal distribution, e.g. Income. I was talking about datat that follows the normal distribution. The bigger the sample the more it fits the normal curve.

I need to read about p-values cause as far as I remember they had little to do with the sample size. I may be wrong of course. : )

[_aj_]

You are wrong. The parameter estimates of the model becomes more accurate and the p-values gets smaller with increasing sample size, assuming there is a true population effect. This is basic inferential statistics. This is why p-values are of little interest in very large samples- practically irrelevantly small effects (associations, differences of means etc) become significant.

Further, although not very important for my argument, what you say about the normal curve is also wrong as not all natural distributions follow a gaussian distribution. For instance event-data, such as number of road accidents in a given time-period, are Poisson-distributed.

NN

Coleman OWNS Yates. Deal with it.

[Man of Steel]

As I wrote: increasing sample size leads to less bias and more precision regarding the estimate of the correlation. Further, it leads to less uncertainty regarding generalizability from sample to population through diminishing p-values. (less chance of both type 1 and type 2 errors).

NN

Nope. A bigger sample means the data will be closer to the normal curve (Central limit theorem) thus when you use it to calculate percentages, probabilities, etc, results will be more accurate. This doesn't translate in lower p-values, for instance, you can have a n = 1000 perfectly fitting the normal curve, but if there's not statistical significance between the things being measured, then P-values aren't going to be on the low side.

Increasing the sample size doesn't result in lower p-values.

You are wrong. The parameter estimates of the model becomes more accurate and the p-values gets smaller with increasing sample size, assuming there is a true population effect. This is basic inferential statistics. This is why p-values are of little interest in very large samples- practically irrelevantly small effects (associations, differences of means etc) become significant.

Further, although not very important for my argument, what you say about the normal curve is also wrong as not all natural distributions follow a gaussian distribution. For instance event-data, such as number of road accidents in a given time-period, are Poisson-distributed.

NN

Sure, not every set of data follows the normal distribution, e.g. Income. I was talking about datat that follows the normal distribution. The bigger the sample the more it fits the normal curve.

I need to read about p-values cause as far as I remember they had little to do with the sample size. I may be wrong of course. : )

And in the end you can throw out all statistics or find "studies" that indicate the opposite of "religion bad" and the VAST majority of the unbelieving world will always.....ALWAYS...... remain firm on the following:

theists = dumb
atheists = smart

The fact that studies have been conducted around this general notion of "religion bad" baffles me because with or without the "studies" virtually no one in the unbelieving community will ever think differently about theists LOL.  It will always be the following:

[Raymondo]

These studies are important because they can help give direction to children and people with underdeveloped critical abilities.

[Man of Steel]

These studies are important because they can help give direction to children and people with underdeveloped critical abilities.

Yes of course, for the betterment of children and underdeveloped thinkers.  

[Raymondo]

Yes of course, for the betterment of children and underdeveloped thinkers.  

Precisely, once a mind is indoctrinated it may never recover completely.

[Man of Steel]

Precisely, once a mind is indoctrinated it may never recover completely.

Thankfully we have your ilk to cascade the light of truth on these poor wretches.

[Raymondo]

Thankfully we have your ilk to cascade the light of truth on these poor wretches.

You act like a sulking child. A sad spectacle, indeed.

[Man of Steel]

You act like a sulking child. A sad spectacle, indeed.

You can insult me if you want.  One more and we're done for today.

[Raymondo]

I personally think that is a terrible question.

There is nothing in the physical world that shows any remote evidence towards a "start" or creation, at all.  That is likely a human concept that will quite possibly eventually have no true place in science.  Everything can be tracked to merely changing form, which can be attributed to general interaction with other bits of "stuff" and the startling implication of "never" an end and the constant flux of interaction indicates that you will absolutely know consciousness again.

There is no evidence of any creation whatsoever.  There is no evidence of the word "nothing" holding true, anywhere.

When you start to contemplate whether our language is in fact leading us to believe in false concepts (ie, creation, nothing etc), things become less paradoxical.

Also, they become highly less scary in terms of this idea that you must grasp to the idea that you will live on, beyond death - because it is 100% true that we know of nothing that is "destroyed" in the act of death, other than the configuration of the components that make you up.

The components you speak of give rise to consciousness. Consciousness, like everything else, is a result of natural processes. When they are destroyed, so is consciousness.

[Man of Steel]

The components you speak of give rise to consciousness. Consciousness, like everything else, is a result of natural processes. When they are destroyed, so is consciousness.

Which natural processes?  How do they interact with one another?  What initiates the processes?

[10pints]

The components you speak of give rise to consciousness. Consciousness, like everything else, is a result of natural processes. When they are destroyed, so is consciousness.

Not a fan of Cartesian duality then...

[funk51]

i'm not an idiot, but i hope that there's a heaven and pray there ain't no hell

[OB1]

i'm not an idiot, but i hope that there's a heaven and pray there ain't no hell

You can't have one without the other I guess.