Intelligence and Religiosity

[JonathanL]

Still dodging I see...

Shuny posts his full name whenever he writes one of his poems here, like this post. I'm taking this to mean that he doesn't mind us referring to him with his full name.

Doing a google search on his name reveals that he's a "semi-retired" soil scientist, and when I search on google for "Frank A. Doonan soil scientist" (without the quotation marks), I get results like http://soilsassociation.org/pdf/Hist...-%203-1974.pdf (Do a control + F search for "Doonan")

http://www.sssnc.org/annual-meetings...roceedings.pdf (Go to the page numbered 78)

http://www.sssnc.org/annual-meetings...roceedings.pdf (Page numbered 13)

There is probably more, but that seems to be the most interesting/relevant results imo.

[seer]

Shuny posts his full name whenever he writes one of his poems here, like this post. I'm taking this to mean that he doesn't mind us referring to him with his full name.

Doing a google search on his name reveals that he's a "semi-retired" soil scientist, and when I search on google for "Frank A. Doonan soil scientist" (without the quotation marks), I get results like http://soilsassociation.org/pdf/Hist...-%203-1974.pdf (Do a control + F search for "Doonan")

http://www.sssnc.org/annual-meetings...roceedings.pdf (Go to the page numbered 78)

http://www.sssnc.org/annual-meetings...roceedings.pdf (Page numbered 13)

There is probably more, but that seems to be the most interesting/relevant results imo.

OK, i'll buy that.

[Jichard]

Kbertsche is a scientist.

Which is as irrelevant as Teal claiming to be a DIS, even though Teal was actively misrepresenting the scientific evidence on the efficiacy of condoms, a point Sam and I had to correct her on, even as you defended her ridiculous claims.

I could not care less what Kbertsche claims to be, if he makes ridiculous claims on the relevant topic.

What are you JerkTard (beyond professional troll)?

You can run along now, troll.

[Jichard]

I'm just a lowly electronic technician. I don't play a scientist on Tweb like Shuny...

You just selectively abuse and prostitute science to justify your own agenda.

Since you ridicule science and describe it as a false God what purpose would it serve to provide my qualifications as a scientist? Would any qualifications presented by by anyone be meaningful to you based on your view of science?

Based on your history these posts only serve as abusive ridicule to further your agenda, and that of your cheerleader gallery.

That's how he operates. He misrepresents science and casts aspersions on those who actually care about scientific evidence. Just another reason that I don't take him seriously and don't think he has a shred of intellectual honesty

Nope, I only cite scientific work because that is your god. You live and die by scientific studies, so I quote your god back to you on occasion.

[One Bad Pig]

Which is irrelevant since the research in question was replicated multiple times using multiple techniques, as you would know if you read the paper.

Jichard,

Since you seem to think you've proved something here...

Please learn the difference between meta-analysis and replication. Also, learn to distinguish between hard and soft science. Thanks.

But I'm glad that you're willing to disregard scientific evidence because you read something in The Guardian.

I'm not, but I recognize that I'm not going to change your mind on that. Studies that use a small sample size combined with selection bias just aren't very convincing IMO - even when you collect a bunch of them.

[Jichard]

Jichard,

Since you seem to think you've proved something here...

Please learn the difference between meta-analysis and replication.

I already know the difference, though you don't seem to.

You do realize that a meta-analysis can be done of multiple studies that have replicated the same findings, right? Multiple studies replicate similar findings, often using different methodologies to yield those findings. And these findings are then summarized and compared in a meta-analysis. So the meta-analysis is actually based on replication. You'd know this if you actually read the scientific research in question before commenting on it. See this for a good introduction to the subject, especially slide 8 on page 2. Or read a scientific paper, such as the following paper which provides a good practical application of meta-analysis in the context of replication:

"Meta-analysis of genetic association studies supports a contribution of common variants to susceptibility to common disease"
http://www.nature.com/ng/journal/v33...ll/ng1071.html

"By contrast, underpowered non-significant studies of real associations with modest genetic effects can reasonably account for much of the variability in replication. Consistent with this model, 8 of the 25 original positive reports were replicated in a meta-analysis of follow-up studies, with modest genetic effects that would be hard to replicate in small studies (Table 2)."

This is all fairly introductory stuff when it comes to statistical analysis. I've done scientific research, have read many meta-analyses, and understand the rationale for doing such meta-analyses. So I know this. Therefore please stop trying to misrepresent science to those who actually do scientific research and read scientific scientific. Because you won't fool those people. Thanks.

Also, learn to distinguish between hard and soft science. Thanks.

Oh, I know they're different. I also know that physics is different from biology. It's just like (unlike you or Kbert) I don't pretend that soft science isn't science nor that soft science lacks reputable scientific evidence. Yes, soft science is often a bit trickier to do than certain hard sciences. But so what? That does nothing to undermine the scientific research produced by soft science, anymore than the fact that immunology is sometimes hard and sometimes messy means there's no evidence within immunology. So please stop making up excuses for why you reject replicated scientific evience that's inconvenient for your pet ideology. Your tried that whole it hasn't been replicated thing, and that failed. Said failure also exposed the fact that you don't seem to grasp the relationship between meta-analysis and replication.

By the way: the above paper I mentioned is a paper in the hard science of genetics. Yet it uses a meta-analysis in the context of replication. So given that you can't use the it's soft science, not hard science excuse for dodging this paper, what will be the excuse you resort to this time, in your attempt to undermine the scientific tool known as "meta-analysis"? Come on; be inventive in your rationalizations. I like seeing denialists try that.

I'm not, but I recognize that I'm not going to change your mind on that.

Actually, you did. You disregarded the scientific evidence discussed in the OP by citing a Guardian article you claimed showed that social scientific results mentioned in the OP hadn't been replicated. Unfortunately for you, your claim failed since the results discussed in the OP were replicated, as you would have known if you'd actually read the studies in question.

Studies that use a small sample size combined with selection bias just aren't very convincing IMO - even when you collect a bunch of them.

OK, now you're again showing that you're not familiar with this subject. OR you're making stuff up. First, you haven't shown any "selection bias" in the research. That's likely just another claim you made up in attempt to dodge addressing the scientific research mentioned in the OP. Second, you haven't shown that the sample size was small. And I know it wasn't small since I (unlike you) actually read the research before commenting on it. Third, I already gave you an example above (from genetics) of how collecting smaller studies together can be helpful in science, given the role of sample size in statistical significance. This is just basic statistics and scientific reasoning: the smaller your sample size, the harder it is to get statistically significant result, even if the underlying trend is real and null hypothesis is not the case. That's why it's useful to get larger sample sizes, including larger samples sizes in the context of a meta-analysis, so that one can get a better understnading of if there really isn't a statistically significant underlying trendo r whether the lack of significance is just due to a small sample size. Hence scientists often use meta-analyses, regardless of whether the topic is a soft science or a hard science.


So yeah, One Bad. Please try not to misrepresent how science works, when talking people familiar with scientific reasoning. You'll probably get caught. Thanks.

[One Bad Pig]

I already know the difference, though you don't seem to.

You do realize that a meta-analysis can be done of multiple studies that have replicated the same findings, right? Multiple studies replicate similar findings, often using different methodologies to yield those findings. And these findings are then summarized and compared in a meta-analysis. So the meta-analysis is actually based on replication. You'd know this if you actually read the scientific research in question before commenting on it. See this for a good introduction to the subject, especially slide 8 on page 2. Or read a scientific paper, such as the following paper which provides a good practical application of meta-analysis in the context of replication:

"Meta-analysis of genetic association studies supports a contribution of common variants to susceptibility to common disease"
http://www.nature.com/ng/journal/v33...ll/ng1071.html

"By contrast, underpowered non-significant studies of real associations with modest genetic effects can reasonably account for much of the variability in replication. Consistent with this model, 8 of the 25 original positive reports were replicated in a meta-analysis of follow-up studies, with modest genetic effects that would be hard to replicate in small studies (Table 2)."

I can't access this because it's behind a paywall, but the quote you provide is not especially flattering, if only 8 out of 25 were successfully replicated. It stands to reason, though, that underpowered non-significant studies would be difficult to replicate.

This is all fairly introductory stuff when it comes to statistical analysis. I've done scientific research, have read many meta-analyses, and understand the rationale for doing such meta-analyses. So I know this. Therefore please stop trying to misrepresent science to those who actually do scientific research and read scientific scientific. Because you won't fool those people. Thanks.



Oh, I know they're different. I also know that physics is different from biology. It's just like (unlike you or Kbert) I don't pretend that soft science isn't science nor that soft science lacks reputable scientific evidence. Yes, soft science is often a bit trickier to do than certain hard sciences. But so what? That does nothing to undermine the scientific research produced by soft science, anymore than the fact that immunology is sometimes hard and sometimes messy means there's no evidence within immunology. So please stop making up excuses for why you reject replicated scientific evience that's inconvenient for your pet ideology. Your tried that whole it hasn't been replicated thing, and that failed. Said failure also exposed the fact that you don't seem to grasp the relationship between meta-analysis and replication.

By the way: the above paper I mentioned is a paper in the hard science of genetics. Yet it uses a meta-analysis in the context of replication. So given that you can't use the it's soft science, not hard science excuse for dodging this paper, what will be the excuse you resort to this time, in your attempt to undermine the scientific tool known as "meta-analysis"? Come on; be inventive in your rationalizations. I like seeing denialists try that.

I know that genetics is a hard science, thanks. Your condescension would be annoying if it weren't so amusing. Genetics is also rather squishy, as far as hard sciences go, because there are many variables we don't understand very well yet. I'm not trying to "attempt to undermine the tool known as "meta-analysis." And it uses a meta-analysis of replicated studies, so I can't help but think that most of your post is an attempt to snow over the fact you were sloppy with your terminology in the earlier post.

Actually, you did. You disregarded the scientific evidence discussed in the OP by citing a Guardian article you claimed showed that social scientific results mentioned in the OP hadn't been replicated. Unfortunately for you, your claim failed since the results discussed in the OP were replicated, as you would have known if you'd actually read the studies in question. .

I never claimed any such thing. Kindly stop misrepresenting me.

OK, now you're again showing that you're not familiar with this subject. OR you're making stuff up. First, you haven't shown any "selection bias" in the research. That's likely just another claim you made up in attempt to dodge addressing the scientific research mentioned in the OP.

Do you understand what selection bias is? One problem with many social science studies is that volunteers are used (self-selection bias).

Second, you haven't shown that the sample size was small. And I know it wasn't small since I (unlike you) actually read the research before commenting on it.

If you'd paid attention, you'd notice that your meta-analysis included studies with sample sizes as small as 20. That's tiny.

Third, I already gave you an example above (from genetics) of how collecting smaller studies together can be helpful in science, given the role of sample size in statistical significance. This is just basic statistics and scientific reasoning: the smaller your sample size, the harder it is to get statistically significant result, even if the underlying trend is real and null hypothesis is not the case. That's why it's useful to get larger sample sizes, including larger samples sizes in the context of a meta-analysis, so that one can get a better understnading of if there really isn't a statistically significant underlying trendo r whether the lack of significance is just due to a small sample size. Hence scientists often use meta-analyses, regardless of whether the topic is a soft science or a hard science.

Sure. A meta-analysis, however, is only as good as its inputs.

So yeah, One Bad. Please try not to misrepresent how science works, when talking people familiar with scientific reasoning. You'll probably get caught. Thanks.

Right back atcha.

[Jichard]

I can't access this because it's behind a paywall,

I would have never thought that a conservative Christian who rejected the evidence-based scientific consensus, would have trouble accessing scientific papers discussing peer-reviewed scientific evidence.

... Oh wait, I'm not surprised by that at all. It's just Dunning-Kruger in action: the scientifically-uninformed making bolder claims than they have any right to.

but the quote you provide is not especially flattering, if only 8 out of 25 were successfully replicated. It stands to reason, though, that underpowered non-significant studies would be difficult to replicate.

What makes you think it's not flattering? Did you understand the quote you read?

The point of the quote is that the small studies in question were examining a real underlying trend (i.e. a trend that was not statistical noise), but that the sample size of the studies (and thus the power of the studies) was unable to yield statistically significant results even given the real trend. That's why a meta-analysis would need to be performed to counter-act the effects of a small sample size.

I know that genetics is a hard science, thanks. Your condescension would be annoying if it weren't so amusing.

I find it hard not to condescend to a conservative Christian who rejects the scientific consensus and makes smug, uninformed comments about science, yet can't even access peer-reviewed scientific evidence. I find it amusing that you can make claims about topics you're clearly uninformed on.

Genetics is also rather squishy, as far as hard sciences go, because there are many variables we don't understand very well yet.

You again don't seem to know what you're talking about.. If you had an even basic understanding of genetics, you'd know that it has a number of tools for controlling for said variables. Did you pay attention in your high school science class when they went over those things called constants and control variables? For example, we have genetic systems in mice, Drosophilia, zebra fish, etc. that are almost completely genetically identical, except for the genetic loci we're interested in investigating. We have studies on identical twins to help control for genetic factors. We have large epidemiological studies that have a lot of power due to their large sample size, and are thus able to help abstract away from the effects of a lot of control variables, and so on.

Knowing those sorts of things is the difference between being informed on science, as opposed to someone who only states empty platitudes like there are many variables we don't understand very well yet as an excuse for not knowing how we control for variables.

Really, you're reasoning is a shallow form of we don't know this yet, since there's other stuff we don't know. It's as silly as someone saying that:

Astronomy is also rather squishy, as far as hard sciences go, because there are many variables we don't understand very well yet. So the we don't know that the Earth is flat, and we shouldn't accept meta-analysesin astronomy

Seriously, do you think there's any science in which we know each and every variable? You could apply your empty platitude to every science, hard or soft.

I'm not trying to "attempt to undermine the tool known as "meta-analysis." And it uses a meta-analysis of replicated studies, so I can't help but think that most of your post is an attempt to snow over the fact you were sloppy with your terminology in the earlier post.

I never even used the term "meta-analysis" in the post you were responding to before. I instead talked about replication. You chose to bring up "meta-analysis" since you saw it in the title of the paper from the OP, and you were too uninformed on the topic to realize that the paper was doing a meta-analysis of replicated work. So really, you were the sloppy, uninformed one:

Which is irrelevant since the research in question was replicated multiple times using multiple techniques, as you would know if you read the paper.

Jichard,

Since you seem to think you've proved something here...

But I'm glad that you're willing to disregard scientific evidence because you read something in The Guardian.

Please learn the difference between meta-analysis and replication. Also, learn to distinguish between hard and soft science. Thanks.

I'm not, but I recognize that I'm not going to change your mind on that. Studies that use a small sample size combined with selection bias just aren't very convincing IMO - even when you collect a bunch of them.

I never claimed any such thing. Kindly stop misrepresenting me.

No, you did:

Further, psychological studies are notoriously difficult to replicate - cognitive studies were shown to have a whopping 50% repeatability rate. A meta-analysis is only as good as the input data (and when the data is this unreliable, it's doubtless not difficult to cherry-pick a dataset that gives the results you want).

Do you understand what selection bias is? One problem with many social science studies is that volunteers are used (self-selection bias).

Apparently you don't know what "selection bias" is. Here's the link I gave you before:

See this for a good introduction to the subject, especially slide 8 on page 2.

Go to slides 11 and 12 on page 3. That should correct your mistake.

Also, you're now making stuff up to cover for the fact that you haven't read the studies in question. If you doubt that, then feel free to show how many of the studies under discussion used volunteers only. You won't be able to answer, because you don't know how many of them did that. Instead, you're just making up reasons to reject research you find inconvenient, without bothering to check if your reasons actually apply to the research in question.

If you'd paid attention, you'd notice that your meta-analysis included studies with sample sizes as small as 20. That's tiny.

Which is irrelevant, since the power of a meta-analysis isn't determined by the sample size of the smallest study. It's instead determined, in large part, by the overall size of the sample across multiple studies that are being compared. That's part of the rationale for a meta-analysis. Seriously, what you're saying is as ridiculous as saying the following:

Yes, I know your study is made up of 400 members per group. But I can point out a smaller population of 20 members in your group. And 20 is tiny. So I'm going to ignore the rest of the 380 members of the sample

You seem to think that statistical power is determined by the smallest sample size you can pick out of the larger whole. That is absurd.

Sure. A meta-analysis, however, is only as good as its inputs.

Ah, now I see the problem more clearly. You seem to think that a meta-analysis is only as powerful as the smallest study included in the meta-analysis. I didn't originally think you thought, since that's such a ludicrous and uninformed claim to believe. Forgive for thinking you were better than that. A meta-analysis is much stronger than the smaller study included in the study; if that was not the case, there would be no point in doing a meta-analysis and scientists would not bother with meta-analyses. Meta-analyses are more powerful for a number of reasons, one of which is that they include more than just one study and therefore have a larger sample size (and thus more statistical power, all other things being equal) than any of those individual studies.