"Race" in College Admissions FAQ & Discussion 3

Applying to College
961

<p>

</p>

<p>It’s already been demonstrated in population samples from (and within) the USA, Europe, East Asia, and Africa. There is a natural clustering of individuals’ genetic profiles derived without any knowledge of social race/ethnic labels, by running standard statistical algorithms on the gene measurements. The output of these algorithms are nothing but “scientific criteria for race or ethnicity”, and the races and ethnic groups they isolate match the pre-existing social constructs in all cases considered so far. </p>

<p>For instance, sorting individuals by their position on the principal axis of genetic variation, gave 100 percent accurate separation of Jews from non-Jews in a mostly American sample of several hundred people; all self-reported Jews were on one side of the scale and the non-Jews on the other side. Two principal components separated Finns and Swedes at 90 percent accuracy (and using more components would of course improve the power of a test to distinguish those groups). Similar separations are observed within East Asian and African populations, where socially-defined “race” or “ethnic group” labels closely match the genetically discovered clusters of gene profiles. Pictures of the samples show a very clear clustering in 2 dimensions, and approximately replicate a geographic map (again, without any geographic input, but purely from the genetics). </p>

<p>

</p>

<p>That’s a straw-man, of course; the content-free terms “separate” and “subspecies” are there for demagogic effect. </p>

<p>DNA studies do indicate that statistically separable groups exist, can be classified, and line up with the socially constructed notions of race and ethnicity. Whether to go so far as to call those groups “subspecies” is a linguistic and political question, not a biological one. The biological fact is that the groupings exist and generally confirm the pre-genetic classifications, in the same way that DNA has solidified the qualitative taxonomy of Linnaeus.</p>

<p>

</p>

<p>Nobody who has read the current literature could claim this. 100 percent and 90 percent accurate separation using one or two principal components is a very strong result; it means that separation of groups by genetics is rather easy. Far easier than anyone expected before seeing the data. </p>

<p>

</p>

<p>That has nothing to do with the ability to objectively cluster individuals’ genomes into racial groups, and suggests a misleading mental picture of “big blobs that mostly overlap” when the published clustering diagrams are of blobs that are far apart or barely touch each other near the edges.</p>

<p>

</p>

<p>The emergence of identifiable groups with visible differences in the last 20000 years or less indicates that human evolutionary divergence is extremely rapid. This supports the notion of racial and ethnic distinctions having a genetic basis beyond the simple visible characteristics. </p>

<p>

</p>

<p>That’s a nice theory, but it’s a load of nonsense. The comparison of “within group” and “between group” variation says nothing about the ability to reliably discriminate, say, Jews from non-Jewish Europeans, Finns from Swedes, and other groups in geographic proximity.</p>

962

<p>Marcus Feldman sums it up best, and he is one of the leading researchers on the issue: knowing the genome of an individual is a very reliable predictor of the individual’s ancestry group, but knowing an individual’s ancestry group is a poor predictor of the individual’s genome. And if “Jew” is a reliably identifiable ancestry group, why isn’t “Jew” one of the checkboxes on federal “race” category forms? That is why I put “race” in quotation marks to indicate the dubiousness of the concept: no national government anywhere on earth refers to scientific information to classify ethnicities for purposes of that government’s data-gathering or preferential policies. The United States Census </p>

<p>[Black</a> or African American persons, percent, 2000](<a href=“http://quickfacts.census.gov/qfd/meta/long_68176.htm]Black”>http://quickfacts.census.gov/qfd/meta/long_68176.htm) </p>

<p>specifically says “These categories are sociopolitical constructs and should not be interpreted as being scientific or anthropological in nature,” and that statement made by professional demographers who know the data should be taken seriously.</p>

963

<p>It’s important too to distinguish genes that best show serial founder effects in human populations and thus distinguish ancestry groups, and genes that result in noticeable differences in phenotype. Many of the genes that show strong founder effects don’t have any known effect on phenotype (showing up only in laboratory tests at the gene level), and indeed may not have any meaningful effect on phenotype, while many of the genes that produce meaningful differences in phenotype are under selection pressure such that similar selection environments operating on differing ancestry groups produce similar frequencies of (differing) genes for a given phenotype. (Example: there are various genes that produce dark skin, and different genes for dark skin in different ancestry groups, but by and large dark skin appears most often as a phenotype among ancestry groups that live in sunny climates, regardless of whether or not the ancestry groups are closely related.) The remarks by Feldman at the end of the question and answer session following his Nobel Conference lecture </p>

<p>[Marcus</a> W. Feldman lecture | Nobel Conference - Gustavus Adolphus College](<a href=“http://gustavus.edu/events/nobelconference/2008/feldman-lecture.php]Marcus”>Marcus W. Feldman lecture | Nobel Conference - 2008) </p>

<p>go into this issue in more detail, of course, and mention some of the problems of misinterpretation that some popular accounts of his research get into.</p>

964

<p>

</p>

<p>Feldman is of course affirming the biological reality of race, and wants a much larger number of biologically valid sub-races added to the vocabulary. </p>

<p>People have been using terms like Middle Eastern, Near Eastern, Semitic, and subdivisions thereof (Jewish, Arab, Persian, Kurdish, etc) for centuries to describe ancestry groups originating in the mideast, so Feldman isn’t asking for anything new. Just a return to the older style of keenly race-conscious discourse popular a century ago, when geography was better studied and colonialism the order of the day.</p>

<p>

</p>

<p>Feldman is not only admitting that race has a biological basis, he is saying that we need more of it. </p>

<p>Thus, whereas we might today refer simply to Welshmen or Jews, Feldman might advocate replacing this with subgroups for northern and southern Welshmen, or throwing out “Jewish” as an ancestry descriptor and replacing it by several subgroups of Jews (Sephardic, Ashkenazic, Mesopotamian, … ). </p>

<p>That’s fine, but Feldman’s position concedes from the outset that race exists; he just wants the language to better approach the resolution of modern genetic analyses. That’s very different from race denialism: the claim that no biological race exists except one human race.</p>

965

<p><a href=“%5Burl=http://talk.collegeconfidential.com/1062790325-post961.html]#961[/url]”>quote</a> It’s already been demonstrated in population samples from (and within) the USA, Europe, East Asia, and Africa. There is a natural clustering of individuals’ genetic profiles derived without any knowledge of social race/ethnic labels, by running standard statistical algorithms on the gene measurements. The output of these algorithms are nothing but “scientific criteria for race or ethnicity”, and the races and ethnic groups they isolate match the pre-existing social constructs in all cases considered so far…

[/quote]
</p>

<p>The argument that:</p>

<p>

</p>

<p>is addressed in ‘[Does</a> ‘Race’ Have a Future?](<a href=“http://stanford.edu/~joelv/teaching/167/kitcher%2008%20-%20the%20future%20of%20race.pdf]Does”>http://stanford.edu/~joelv/teaching/167/kitcher%2008%20-%20the%20future%20of%20race.pdf)’:</p>

<p>

</p>

<p>The statistical “scientific criteria for race or ethnicity” being argued for is influenced by the [a</a> priori](<a href=“http://en.wikipedia.org/wiki/A_priori_(statistics)]a”>Prior probability - Wikipedia) groupings imposed on the analysis. </p>

<p>Which then gets to the heart of this threads discussion, ‘[…What</a> Difference Makes a Difference?](<a href=“http://www.itvs.org/outreach/race/RaceWhatDifference.html]...What”>http://www.itvs.org/outreach/race/RaceWhatDifference.html)’:</p>

<p>

</p>

<p>By definition, everyone is different. The clustering studies themselves detail the rich genetic diversity within clusters. Even genetic twins have different [url=<a href=“http://www.pbs.org/wgbh/nova/sciencenow/3411/02.html]epigenetic[/url”>Epigenetics | NOVA | PBS]epigenetic[/url</a>] responses to their enviroments. </p>

<p>Any argument for the social policy of ‘race’-based college admissions based on statistical groupings – that are influenced by the original selection of those groupings-- should prove its utility over remedies for the social assistance of individuals that simply involve income and asset tests. Affirmative action based on a individual’s participation in society can be justified without resorting to ‘race’. The social sciences inform us of the perils of highly [url=<a href=“http://en.wikipedia.org/wiki/Social_stratification]stratified”>Social stratification - Wikipedia]stratified</a> societies<a href=“e.g.:%20the%20%5Burl=http://www.cyberessays.com/History/107.htm]French%20Revolution[/url]”>/url</a>.</p>

966

<p>StitchInTime is correct that the most meaningful differences among human beings that are influenced by genes occur within rather than between ancestry groups. And my point continues to be that no national government anywhere on earth defines “race” or “ethnicity” in any manner but by political convenience, often contrary to the minimal amount of science that is available on the subject. Biology is not the answer for how college applicants should fill out their ethnic identification questionnaires, if they choose not to exercise their right to decline to answer the questions at all. </p>

<p>The correct statement of the science, per Feldman and other researchers on the topic, is </p>

<p>[Marcus</a> W. Feldman lecture | Nobel Conference - Gustavus Adolphus College](<a href=“http://gustavus.edu/events/nobelconference/2008/feldman-lecture.php]Marcus”>Marcus W. Feldman lecture | Nobel Conference - 2008) </p>

<p>

</p>

967

<p>Just as a reminder, I’ll be updating the FAQ and posting a new thread once the new application season’s application forms have been posted. Good luck to all the high school class of '10 applicants.</p>

968

<p>I know this topic has been discussed before on various threads, but I’m having trouble locating just where they are, so I thought I’d just make this thread to ask my question.</p>

<p>I know we aren’t forced to fill out our ethnicity on the common app (or any application at all for that matter). Moreover, as I understand it, ethnicity is only supposed to help your chances – not hurt them. However, that seems to merely be a statement of political correctness. Being Asian (Chinese to be more specific), I can’t seem to shake off that uneasy feeling that filling in the “Asian” bubble would actually work against me for HYPSM and other Ivies simply because there are so many Asians at the top universities now. </p>

<p>I was leaning towards simply leaving the ethnicity section blank, but I don’t know how effective that will be because my last name, which would show up on my transcript, is a fairly common Chinese surname and would probably give away my ethnicity anyway.</p>

<p>Simply put, will telling HYPSM I’m Asian on my common app work against me? Or would leaving it blank give the wrong image to the adcoms? Would it even be effective to leave the section blank given my last name?</p>

969

<p>I don’t think leaving the ethnicity section blank on the application would really hurt your chances in any way. The admission officers go through too many applications to make cynical assumptions about prospective students.</p>

970

<p>

</p>

<p>Your question has been merged into the main FAQ thread. [Lots</a> of colleges admit lots of applicants who don’t self-report ethnicity](<a href=“http://talk.collegeconfidential.com/1061808252-post4.html]Lots”>http://talk.collegeconfidential.com/1061808252-post4.html), which is every applicant’s legal right, so this shouldn’t be a problem.</p>

971

<p>What is your ethnicity? If you are part hispanic then you can check the box. I don’t think where your uncle and Dad lived factors into it. Yes, they check. Race is a factor in who gets admitted and who doesn’t so they check. They don’t want applicants to be claiming to be a minority when they aren’t.</p>

<p>I’m confused with your question about aid. I’ve heard the same thing, that you are better off not applying for financial aid. But if you need aid then what are you going to do? If you don’t need aid then there is no point in applying for it anyway because you won’t get any. I also think that whether applying for financial aid hurts your chances of being accepted is more of a factor at smaller schools that rely more on tuition for their operating costs. I personally believe that Harvard, Yale and Princeton are still need blind.</p>

972

<p>Well, are you actually Hispanic? Having a relative live in a foreign area does not change that relative’s ethnicity…</p>

<p>EDIT: Originally Middle Eastern? Then no, you are absolutely not Hispanic.</p>

973

<p>I don’t think that ethnicity is something that you can really be confused about. If your uncle and father lived in Puerto Rico but are not of Hispanic descent, then you are not Hispanic. A Japanese person living in Puerto Rico is still considered Asian. If your family is Middle Eastern, then you are Middle Eastern. Your race is about blood, not residency.</p>

<p>Applying for financial aid will have no effect at a school that considers itself need-blind, but may hurt your chances at schools that are not.</p>

974

<p>

</p>

<p>That essay doesn’t come close to addressing anything. It is a reflection of the author’s desire to claim to have refuted racialistic thinking, without actually knocking down the biology and statistics involved.</p>

<p>

</p>

<p>It’s been known for ages, and in recent times confirmed in all sorts of medical and biological studies, that socially labelled “race” correlates with a long list of things deemed to be important: strength, longevity, height, temperament, masculinity, disease susceptibility, etc. Children often make racial distinctions easily and naturally, without social priming (they also fail to notice some racial distinctions, in the absence of given social labelling; everyone these days probably knows anecdotes of young children “not seeing skin color”). Adults who have dated, lived, or worked with many people from different racial groups tend to have their own bank of observations. It is brainwashing to try and pre-empt any sort of racial cognition by demanding a prior justification (“defense” as the author puts it) for thinking that recognizes some degree of natural population grouping into visible genetic substrata. </p>

<p>

</p>

<p>That claim doesn’t fit well with the fact that chimps share 90 to 98 percent of their DNA with humans, depending on how you count. It’s also a false picture for the mathematics, as I discussed in the earlier posting. There need not be any overlap between groups whose “within group” variation predominates over “between groups” variation. </p>

<p>

</p>

<p>Another mistake. The clusters are differently heterogeneous. Not only do the clusters have little or no overlap (races can be distinguished) but the shapes and sizes of the clusters are different (traits vary differently within races). This is very visible in the PCA plots from genetic studies over the past several decades. </p>

<p>

</p>

<p>Completely false. First of all, the PCA studies don’t assume or impose any grouping at the outset. They just reveal structure that happens to align with the socially defined labelling (race, tribe, nation, family). One can go ahead and run a grouping algorithm to objectively label the picture instead of declaring based on human judgement what the clusters should be, but it makes little difference because the result is generally the same. </p>

<p>Second, there are grouping algorithms that don’t assume anything in advance about the number or nature of the groups. When using algorithms that require a given number of clusters or other exogenous inputs, no scientist would simply assume a fixed number; these studies run the algorithm with different specifications. Feldman’s papers do this, for example. The outputs from the different specifications are consistent with each other. That all approaches point to the same thing indicates that there is a real grouping structure that all methods are discovering independently. And this grouping is essentially the same as the social labelling of race, ethnic, and tribal groups. </p>

<p>

</p>

<p>A huge mistake. The MSTR’s are involved in regulating genes, and may account for most of the evolutionary difference between humans and other animals. Besides, it is well understood that even relatively small samples of observable characteristics, be they phenotypes or particular DNA markers, are sufficient to detect the statistical signature of clustering when it is present. Whether the observations are of coding or non-coding DNA or anything else genetically based is immaterial; they only need to vary in correlated ways. That’s the reason the social and genetic race classifications agree: if you take these 300 junk-DNA characteristics or the dozens of visible phenotypic traits involved in the social labelling, either one is sufficient to start zooming in on the group structure that exists, if it exists.</p>

975

<p>

</p>

<p>That’s a revision of your actual and often repeated point, which has been to deny (and indeed to suppress, such as on college applications) social race labelling as being an anachronistic delusion. You further claimed that this is not only your opinion, but is the authoritative scientific consensus. In reality, science truthfully if uncomfortably supports the racial distinctions, however anachronistic they may be and however troubled the historical precedents. </p>

<p>

</p>

<p>That’s an interesting point but is logically separate from your implications that “race” should always be written between quotation marks. </p>

<p>

</p>

<p>It is not, standing out of context, a correct statement of the science. </p>

<p>

</p>

<p>Feldman did not show any destruction of race “typology”, that is, the use of a combination of phenotype, ancestry and geographic-origin information to socially attribute race, ethnic, tribal or other quasi-genetic classification to individuals.</p>

<p>Feldman did show that an exclusively geographic typology is refuted by genetics, but that isn’t what anyone means by “race”.</p>

<p>

</p>

<p>That’s the standard mistake. What is of interest is the simultaneous distribution of multiple genes whose presence or frequency are correlated. Evolution may enforce several related traits, for example, so that a population with a high rate of one also has high levels of the others. This leads to drastic differences between groups that can be all-or-nothing (pigmentation, hair type). In the paper reported in Feldman’s lecture, out of about 2000 markers, about a thousand had a frequency of zero in some groups. </p>

<p>

</p>

<p>One can infer correlations between observed traits and a slew of unobserved ones. Asian continental ancestry correlates with lactose intolerance; maybe one shouldn’t serve hot cocoa at the gathering for visiting Chinese students? Given more racial information (are they Tibetans or Mongolians?) the opposite holds. You can’t always give out a questionnaire to determine who can digest what.</p>

976

<p>This reminds me of an article in the news that I read awhile back. It seems there was a white guy who was born in, raised in, spent his entire life in, and just recently moved to the US from Africa. His college professor stirred up a hornet’s nest when he insisted that the students discuss their ethnicity.<br>
Well, this guy said he was African-American, which he truly was, but a black girl in the class got offended and yelled at him, telling him he was insulting her. Everyone in the class, including the prof, demanded that he apologize to this black girl who’d never even BEEN to Africa, nor had her immediate family. He refused to apologize (I think), and he got into trouble at his college as a result. Don’t remember how it all ended.</p>

977

<p>**** these classifications. I hate collectivism. We should look at each other as individuals, not as members of subsections of the human race.</p>

978

<p>

</p>

<p>You can follow the link to the federal definitions from one of the FAQ posts on this thread ( [post</a> #2](<a href=“http://talk.collegeconfidential.com/1061808213-post2.html]post”>http://talk.collegeconfidential.com/1061808213-post2.html) ) and see if you fit the rather vague definition of Hispanic ethnicity. </p>

<p>Or you can choose to mark nothing at all, if that is what you would rather do. </p>

<p>In general, colleges recently claim to want to admit more low-income applicants, and to be ready to use endowment funds to subsidize their educations. The way to test whether or not that is true is to apply.</p>

979

<p>@ [#976[/url</a>]</p>

<p>Assume for the moment that one were to accept your argument that [url=<a href="http://www.sciencemag.org/cgi/content/abstract/298/5602/2381]Feldman’s"&gt;http://www.sciencemag.org/cgi/content/abstract/298/5602/2381]Feldman’s&lt;/a&gt; genetic clusters](<a href=“http://talk.collegeconfidential.com/1062819536-post976.html]#976[/url”>http://talk.collegeconfidential.com/1062819536-post976.html) are essentially “race” clusters as well. For sake of argument, let’s say the clusters are A, B, C, D , E and F corresponding with Feldman’s “six main genetic clusters”. How much “race” inter-breeding would have to occur (A with B, A with C, A with D, A with E, B with C, etc…) to remove the salient, genetically-based “race” differences you’re arguing for? </p>

<p>Since you disagree with the arguments in ‘[Race</a> and Gene Studies: What Difference Makes a Difference?](<a href=“http://www.itvs.org/outreach/race/RaceWhatDifference.html]Race”>http://www.itvs.org/outreach/race/RaceWhatDifference.html)’, is there any amount of genetic material mixing within our species that would eliminate a need for your justification of “race”?</p>

980

<p>

</p>

<p>thats really stretching it. will you feel good that you got in for lying? I mean its your uncle… not even your dad, but your uncle…by your dad being somewhat the same, im guessing you mean he stayed there for vacation.</p>