For a long time I’ve been noticing (as I’m sure other contributors have) curious parallels between religious creationists (Everything is the work of God! Who lives in the sky!) and left-liberal “culturists” (All group differences are due to Culture! Which drops from the sky!)
The blogger who calls himself One Standard Deviation (not to be confused with the different blogger who calls himself Half Sigma) has done a very neat tabulation of these parallels.
The parallels are interesting, but suppose we identify specific alleles that are associated with higher intelligence and occur in some ethnic groups more than others. That may be academically interesting, but does it have any practical implications? We wouldn’t ignore the work of Gauss just because he wasn’t Chinese.
Pingback: Liberal and religious creationism
Even if, Tony, your hypothetical finding has limited implications for how we treat groups, it could change the way we treat individuals.
Maybe some children should be left behind? Maybe IQ tests do have value in predicting fire-fighting abilities? Maybe it would be better for society to target more of our education budget at the leading edge of the bell curve rather than the trailing edge?
Bringing race (or any grouping) into just clouds the issue.
BTW I have a hunch that the culturist-left would have less of a knee jerk reaction to the science around IQ and heritability if we could divorce the discussion of it from discussions about race.
I say this, not to condemn Tony for focusing on the racial aspect of Bradlaugh’s observation nor to condemn Bradlaugh for mentioning group differences in the first place. Rather, I say it out of despair that people on the left can’t see the science of heritability through the fog of race. I am happy to mock them as much as the next man, but I’d far rather find a way to educate them about the science.
I read The Bell Curve last year and after chapter after chapter of well-reasoned arguments on the data supporting the reality of IQ, its heritability and its consequences for society and public policy, I got to the chapter on race and my heart sank. Whether that chapter had merit or not, it pretty much guaranteed that the rest of the book would be dismissed out of hand. The book – and the discussion – would have been better off without it.
On the contrary, I’m arguing that knowledge about group differences should have little if any bearing on how we treat individuals. For example, we know that Black men are at relatively higher risk from prostate cancer than White men. Should we therefore treat a man for prostate cancer just because he’s Black? Of course not. We only treat him if he as an individual expresses the cancer phenotype. Should we not screen a White man for prostate cancer just because he has less risk than a Black man? Again, no. Depending on other factors, he is still at sufficiently elevated risk to warrant screening.
So it should be with intelligence.
As for treating groups, I should think that the conservative approach is not to treat groups at all, but only to treat individuals.
I think we had a misfire somewhere because I agree with your last post entirely especially your final point.
To recap my argument:
1) Accepting that intelligence exists and is heritable could result in dramatic changes in how we treat individuals.
2) Let’s leave groups out of this as it confuses point 1).
3) Especially racial groups.
1) Accepting that intelligence exists and is heritable could result in dramatic changes in how we treat individuals.
2) Let’s leave groups out of this as it confuses point 1).
3) Especially racial groups.
This is difficult to do as we operate in a public policy sphere which makes no account of group differences in intelligence and fosters a climate of victimization and directs blame at one “group” for the status of other “groups.”
If you’re advocating a policy of individual assessment, then you undercut the whole Liberal view of the public policy sphere, which is entirely predicated upon group level metrics.
> If you’re advocating a policy of individual assessment, then you undercut the whole Liberal view of the public policy sphere, which is entirely predicated upon group level metrics.
I hope so.
I’d like a liberal view of the public policy sphere that is based on sound science.
So there’s a blogger called Half Sigma and another one called One Standard Deviation! I think I’ll start a blog called “One and a Half More Sigmas”.
Oh, yeah, the table was great.
I hope so.
I’d like a liberal view of the public policy sphere that is based on sound science.
This is a very tough battle. If you set out to strip your opponent of everything that they stand for and the principal method they use for forming their coalition, then you’re engaging in a war to the death, rather than just political sparring or bounded battles. When you engage in such a war you shouldn’t willingly disarm. If the left frames everything in terms of group differences then an effective tool to use against them, I believe, are arguments that make them uncomfortable because they are either framed in group terms or undercut groups as the unit of measure. Here is an example of what I mean.
I don’t believe that the left ‘stands for” group level metrics any more than the right ‘stands for’ a religiously-inspired social conservative agenda. It just appears that way when you cherry pick your opponents’ arguments.
I don’t believe that the left ’stands for” group level metrics.
I constantly see the following type of issues in play in Democratic circles:
racial crime disparity;
racial wage disparity;
gender wage disparity;
gender sports disparity;
gender science employment/study disparity;
There is usually more disparity within each of the groups than there is between the group and the designated target of comparison, usually Caucasians and men. The most common policy response from Democrats is “we’ll do something about this disparity that impacts your group.” They seek to develop a dialog/dependency between policy/politician and group membership. If you’re Black, then Democrats have the policies that help you. If you’re Hispanic, then the Democrats have the policies that help you. If there is greater income inequality within the category of “Black Men” than there is between Black Men and White Men, why is the Democratic focus on the smaller problem, racial wage disparity? This isn’t cherry picking arguments, it’s detailing the modus operandi of a Democratic Party view of the world. It works for them.
I am citing:
“Tony: On the contrary, I’m arguing that knowledge about group differences should have little if any bearing on how we treat individuals.”
Men and women definitely constitute two groups. Should medical doctors equally prescribe them testosterone ? Should medical doctors equally screen them for mammalian cancer ? For cervical cancer ? For prostate cancer ?
As for attempts to eliminate the question of groups and races from IQ, see e.g. http://johnderbyshire.com/Opinions/HumanSciences/raceiq.html
, which convincingly shows impossibility to do that.
Your truly, Florida resident.
@Florida resident
I was only referring to racial groups. The argument doesn’t generalize to other kinds of groups (e.g. species, genus, etc). If you have an example in medicine where knowledge about racial groups conveys incremental diagnostic value, I’m all ears.
I am citing: Kevin Lawrence: …
“I read The Bell Curve last year and after chapter after chapter of well-reasoned arguments on the data supporting the reality of IQ, its heritability and its consequences for society and public policy, I got to the chapter on race and my heart sank. Whether that chapter had merit or not, it pretty much guaranteed that the rest of the book would be dismissed out of hand. The book – and the discussion – would have been better off without it. …”
1) Esteemed Mr. Lawrence ! What took you so long to start reading the book in question? Why did you wait from 1994 (when ‘The Bell Curve’ was published) till 2008 ?
2) Citing http://johnderbyshire.com/Opinions/Straggler/073.html
, “Tragic truth, or festive falsehood? The will needs a goal, said Nietzsche, er braucht ein Ziel, and we’d rather will Nothingness than not will, rather believe than know. Nietzsche was right, too.”
Your truly, Florida resident.
Esteemed Tony !
Here they are:
http://ndt.oxfordjournals.org/cgi/content/abstract/6/12/931
http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6T18-3X9S2XT-V&_user=2139851&_rdoc=1&_fmt=&_orig=search&_sort=d&_docanchor=&view=c&_searchStrId=958620867&_rerunOrigin=google&_acct=C000054275&_version=1&_urlVersion=0&_userid=2139851&md5=3825f38fe1ea0578127ffd06077b99ea
http://www.citeulike.org/group/992/article/3499846
And many, many others.
Your truly, F.r.
Esteemed Tony !
I am not sure if 2nd reference was user-specific; here is the regular reference:
Journal of the American College of Cardiology
Volume 34, Issue 3, September 1999, Pages 787-794
Clinical Studies
Racial differences in the significance of coronary calcium in asymptomatic black and white subjects with coronary risk factors*1
Terence M. Doherty BAa, b, Weiyi Tang MDa, b and Robert C. Detrano MD, PhD, FACC, , a, b
And many, many others.
Your truly, F.r.
@Florida resident
The study you cite shows that the risk for coronary disease is different between Blacks and Whites. That there are racial differences in risk factors for various diseases is not in dispute (as is clear from the prostate cancer example that I cited). The point is that racial information doesn’t assist with diagnosis in individual patient encounters.
Dear Tony ! I am not a medic. But judging for the fact that I had to provide those refernces to you, you are neither.
Not being a medic, how do you know that “racial information doesn’t assist with diagnosis in individual patient encounters” ?
I suspedct strongly that it does.
How about info about group category of ethnicity: do you have to perform rather expensive tests for symptoms of Tay-Sachs disease ( http://en.wikipedia.org/wiki/Tay-Sachs_disease ) without knowledge of ethnic origins ?
Your truly, F.r.
For Tony: looks like a post with references still waits moderation. Duplicate:
http://ndt.oxfordjournals.org/cgi/content/abstract/6/12/931
http://www.citeulike.org/group/992/article/3499846
Your F.r.
For Tony:
Nephrol Dial Transplant (1991) 6: 931-935
© 1991 European Renal Association-European Dialysis and Transplant Association
Racial Origin and Primary Renal Diagnosis in 771 Patients with End-Stage Renal Disease
M. Pazianas1, J. B. Eastwood2, K. D. MacRae1 and M. E. Phillips1
1 Charing Cross and Westminster Medical School, Charing Cross Hospital London 2 Division of Biochemical Medicine, St George’s Hospital Medical School London, UK
Your F.r.
@Florida resident
Since you ask, I’m not a medic, but I’m a physician.
You test for Tay-Sachs disease if a patient presents with symptoms of Tay-Sachs disease.
From the Wiki article you cite, “But Tay-Sachs disease is sometimes misdiagnosed at first, because clinicians are not aware that it is not exclusively a Jewish disease.”
All interesting information, but it doesn’t help in individual patient encounters.
Esteemed Tony ! So, you acknowledge that, according to Wikipedia, a lot of practicing physicians _do_ use group characteristic: ethnicity (in this particular case, Jewiish) as a component of diagnosis process.
Do you suggest to forbid this practice ?
Your F.r.
To Tony: you asked for “where knowledge about racial groups conveys _incremental_ diagnostic value, I’m all ears.”
When you are given examlpes not of a decisive, but _incremental_ value, where are your ears ?
@Florida resident
I didn’t say that it isn’t used; I only said that it doesn’t help.
In practice, the clinican isn’t blind to racial information, but as the Wiki article suggests, that information can mislead.
“Incremental diagnostic value” means that the information significantly changes the probability that a disease is present given a set of symptoms, signs, and test results. If a patient presents with an MI, knowing that the patient is Black and therefore has different risk factors from a White adds no incremental diagnostic value in determining that he has an MI.
To Tony: “can mislead”.
But can also lead in right direction.
Otherwise you would have said “always misleads”.
Dict.org :
incremental
adj : increasing gradually by regular degrees or additions;
“lecturers enjoy…steady incremental growth in salary”
Incremental \In`cre*men”tal\, a. (Biol.)
Pertaining to, or resulting from, the process of growth; as,
the incremental lines in the dentine of teeth.
[1913 Webster]
No “significantly” in definition of “incremental”.
If you have an example in medicine where knowledge about racial groups conveys incremental diagnostic value, I’m all ears.
How about this:
False dilemma. It generally has no impact one way or the other.
@TangoMan
Reasonable question, but the diagnosis wasn’t made using information about race. The subjects had symptoms of long QT syndrome, and the diagnosis was botched by using a genetic test that was standardized in Whites. The case report shows the weakness of using genetic testing to confirm a diagnosis. It’s worth noting that the symptoms remained paramount.
Reasonable question, but the diagnosis wasn’t made using information about race.
The example demonstrates that if the diagnosis had factored in relevant information that pertains to racial differences, then incremental diagnostic value would have been attained. The logical consequences that arise from an analysis show that incremental value is added in the diagnostic phase when racial data is included in the analysis. Consider:
Race neutral test is applied to the sons = misdiagnosis.
Race specific test is applied to the sons = proper diagnosis.
Added information pertaining to racial factors improves the diagnostic predictability of the tests, thereby incrementally improving the diagnostic phase of the physician’s treatment.
In light of how the comments have veered into race-based medicine, here is Dr. Sally Satel’s essay on “I am a racially profiling doctor.”
Dear TangoMan !
Thank you for your posts, and for the reference to Dr. Sally Satel’s essay.
Your sincerely, Florida resident.
The example you cite apparently refers to a 2004 paper by Sharma et al. The findings in the paper undermine the claim that the so called race neutral test has any diagnostic value at all since the mutation didn’t cause conduction abnormalities in vitro or disease in healthy Chinese controls. The authors didn’t determine whether the “race specific” test is race specific because they didn’t test Whites or Blacks with long QT syndrome for the presence of the mutation.
In order to show that a test has incremental diagnostic value, you have to demonstrate that the post test probability is statistically different from the pre test probability. The authors didn’t do that with regard to the genetic test. In fact, when the first genetic test didn’t match the phenotypic presentation of the subjects, the test was rejected. A new test was devised that agreed with the phenotypic presentation. The new test wasn’t investigated prospectively for diagnostic value. You can’t show additive diagnostic value for a test that only recapitulates what you already know from the intial presentation.
The new test wasn’t investigated prospectively for diagnostic value. You can’t show additive diagnostic value for a test that only recapitulates what you already know from the intial presentation.
Medical studies are always limited in scope. Your criticisms of the Sharma et al. conclusions are partially addressed by Yan and Wu, 2009:
The Yan study that you cite refers to a different mutation and a different disease (LQT1 rather than LQT2). It has nothing to do with the Sharma study.
@TangoMan
And yes, medical studies are limited in scope, which is why the conclusions need to be limited in scope.
And yes, medical studies are limited in scope, which is why the conclusions need to be limited in scope.
The authors didn’t determine whether the “race specific” test is race specific because they didn’t test Whites or Blacks with long QT syndrome for the presence of the mutation.
They didn’t meet your specified criterion, but they did report:
Another study finds:
These situations presents some useful knowledge, but it’s incomplete, thus creating an opportunity for another researcher to conduct an investigation which specifically addresses your point. We are finding populations with mutations. This is not an insignificant finding.
Where I jumped into conversation was in response to this comment that you had made: “If you have an example in medicine where knowledge about racial groups conveys incremental diagnostic value, I’m all ears.”
Even with the picture not being complete, the knowledge that’s been published in disparate papers does provide a path to incremental diagnostic value. Recall, that the sons of the patient, in my original reference, were misdiagnosed by standard diagnostic procedure and then were properly diagnosed when the findings from a racially focused investigation were applied to their case. I can’t see how a process that corrects misdiagnosis to proper diagnosis doesn’t meet your original condition and why you aren’t all ears? 🙂
As our conversation progressed you’ve moved the goal posts by more restrictively defining what you meant by incremental diagnostic value, and your criticisms of the original reference are predicated upon this more narrow definition of incremental diagnostic value. This iteration of the debate is now focused on professional standards and definitions rather than on whether the sons of the original patient were aided by medical practice that was informed by racially focused science. This is all well and good, but these are two different standards, and perhaps even two different debates.
In other words, I think we’re speaking past each other. I’m referencing studies that seem to confirm a hypothesis of racial disparity in some medical conditions, and I’m assuming that given more time and more study, that these disparate findings, which still possess applicability in medical treatment today, will coalesce into more robust studies which directly address the issues that must be satisfied by the stringent requirements associated with refining medical diagnostic protocols.
Bravo, TangoMan ! With my so-so knowledge of the teminology “incremental”, I could not reach the moving target. You did it.
TangoMan: “These situations presents some useful knowledge, but it’s incomplete, thus creating an opportunity for another researcher to conduct an investigation which specifically addresses your point. We are finding populations with mutations. This is not an insignificant finding.”
Me: I don’t disagree that the findings are significant. I endorse the research, but incremental diagnostic value hasn’t yet been demonstrated.
TangoMan: “Even with the picture not being complete, the knowledge that’s been published in disparate papers does provide a path to incremental diagnostic value.”
Me: Good. Let me know when we get there.
TangoMan: “Recall, that the sons of the patient, in my original reference, were misdiagnosed by standard diagnostic procedure and then were properly diagnosed when the findings from a racially focused investigation were applied to their case.”
Me: The older son was misdiagnosed using a test that has questionable diagnostic value for members of any race, according to the results of Sharma et al. See earlier post. As I recall, the younger son was asymptomatic and had a normal EKG. We don’t know whether or not the new test resulted in a correct diagnosis for the younger son. The presence of the mutation does not necessarily mean that disease is present. Sharma et al point out in the background section of their paper that various mutations don’t always cause disease. How can we know whether or not this newly discovered mutation always causes disease?
Even if the new test did result in a correct diagnosis for the younger son, the same test might also result in a correct diagnosis for members of other racial groups. The authors didn’t apply the test in an ethnically unmatched, disease matched population. Note that the White and Black subjects used in the study did not have the disease. For all we know, some Whites and Blacks with the disease might also have the mutation.
TangoMan: “As our conversation progressed you’ve moved the goal posts by more restrictively defining what you meant by incremental diagnostic value,…”
Me: I’ve been very consistent with my use of that phrase, since I’ve known all along what it means, unlike some other posters around here. The goal post isn’t moving; the kicker is learning where the goal post is. From the beginning, I acknowledged that there are different genotypic prevalences among different races. My point all along is that information about race doesn’t improve diagnosis in individual patient encounters. That may change someday, but there are theoretical reasons to think that if it ever does change, it will be applicable in a relatively small number of diseases. The reason is that phenotype has more predictive power than genotype. That’s because of reduced penetrance, epigenetic modifications, complex interactions between transcription pathways, things like that. Consequently, it’s very difficult but not impossible to show that genotype has incremental diagnostic value. Race is at best an imperfect surrogate for genoytpe; it’s even less likely that the imperfect surrogate will add diagnostic value.
TangoMan: “I think we’re speaking past each other. I’m referencing studies that seem to confirm a hypothesis of racial disparity in some medical conditions,…”
Me: If we are talking past each other, that’s not my fault. I’ve never questioned the hypothesis to which you refer.
TangoMan: “…and I’m assuming that given more time and more study, that these disparate findings, which still possess applicability in medical treatment today,…”
Me: Not yet they don’t
TangoMan: “…will coalesce into more robust studies which directly address the issues that must be satisfied by the stringent requirements associated with refining medical diagnostic protocols.”
Me: Let me know when that happens.