| Topic: |
Science > Philosophy |
| User: |
"" |
| Date: |
21 Sep 2005 05:22:38 PM |
| Object: |
Affirmative action at 3 univeresities: a report |
http://www.ceousa.org/pdfs/VAS%20Report.pdf (you need Adobe Reader to
view it)
"The odds ratio for blacks compared to whites at NCS is 13 to 1, but at
UVA it is 106 to 1 and at William &Mary 267 to 1. In other words, at
UVA the odds of a black student being admitted is more than 100 times
the odds of admission of a white student with the same qualifications.
The odds of admitting a black applicant at William & Mary is more than
250 times the odds of admitting an equally-qualified white applicant.
The odds ratios for Asians at all three schools are less than one,
meaning that Asians are less likely to be admitted than
equally-qualified whites (the odds ratio for Asians at UVA is not
statistically significant). The odds ratios for Hispanics are 2.8 and
1.9 at UVA and NCS, respectively, but less than one at W&M. This means
that Hispanics are somewhat more likely than whites to be admitted at
UVA and NCS but less likely than whites to be admitted at W&M Law (but
this is not statistically significant)."
University of Virginia (UVA) undergraduates, North Carolina State (NCS)
undergraduates, and William and Mary Law (W&M Law) School.
.
|
|
| User: "Chance Hopkins" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
21 Sep 2005 07:32:02 PM |
|
|
<richasiankid@hotmail.com> wrote in message
news:1127341358.848521.111510@z14g2000cwz.googlegroups.com...
http://www.ceousa.org/pdfs/VAS%20Report.pdf (you need Adobe Reader to
view it)
"The odds ratio for blacks compared to whites at NCS is 13 to 1, but at
UVA it is 106 to 1 and at William &Mary 267 to 1.
------------
In other words, at
UVA the odds of a black student being admitted is more than 100 times
the odds of admission of a white student with the same qualifications.
Did they compare the student base against applicants to reach this
conclusion?
Or did they just assume that an equal number of black and whites with the
"same qualifications" applied?
------------
The odds of admitting a black applicant at William & Mary is more than
250 times the odds of admitting an equally-qualified white applicant.
The odds ratios for Asians at all three schools are less than one,
meaning that Asians are less likely to be admitted than
equally-qualified whites (the odds ratio for Asians at UVA is not
statistically significant). The odds ratios for Hispanics are 2.8 and
1.9 at UVA and NCS, respectively, but less than one at W&M. This means
that Hispanics are somewhat more likely than whites to be admitted at
UVA and NCS but less likely than whites to be admitted at W&M Law (but
this is not statistically significant)."
University of Virginia (UVA) undergraduates, North Carolina State (NCS)
undergraduates, and William and Mary Law (W&M Law) School.
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
24 Sep 2005 02:46:14 PM |
|
|
The question they have, I think, is whether groups with same/similar
GPAs and SAT scores get admitted at equal probabilities. I don't think
you can generalize these results into high school student abilities. Is
that what you meant?
.
|
|
|
| User: "Chance Hopkins" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
25 Sep 2005 02:28:55 AM |
|
|
<richasiankid@hotmail.com> wrote in message
news:1127591174.042204.306680@g14g2000cwa.googlegroups.com...
The question they have, I think, is whether groups with same/similar
GPAs and SAT scores get admitted at equal probabilities. I don't think
you can generalize these results into high school student abilities. Is
that what you meant?
It said:
"The odds ratio for blacks compared to whites at NCS is 13 to 1, but at
UVA it is 106 to 1 and at William &Mary 267 to 1.
------------
In other words, at
UVA the odds of a black student being admitted is more than 100 times
the odds of admission of a white student with the same qualifications.
I'm wondering, what was the ratio of applicants? Their conclusion "odds of a
black student being admitted" is in question. I'm saying, what where the
ratio of applicants?
It's hard to discuss "odds of being admitted" without comparing the ratio of
accepted to the ratio of applicants.
Was the ratio of applicants different at each school? Did more blacks apply
at NCS and less at UVA? Is that the reason for the different ratios of
admission?
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
27 Sep 2005 05:02:24 PM |
|
|
I missed your reply hence the delay:
Yes it's stated - see chart 1
"Chart 1 shows the percentage of applicants who were admitted broken
down by
race/ethnic ity without regard to any other characteristic. At UVA,
nearly 65% of all black
applicants were admitted in the Fall of 2003 compared to 36 to 38% of
all other groups. Overall, only 39% of all applicants were admitted. Of
those with race coded, blacks represent about 12% of admissions, Asians
12%, Hispanics 4%, and whites 72%. NCS is considerably less selective
than UVA, with an overall admission rate of 62%. In addition, the
chance of admissions do not differ as much by race as at UVA, and in
fact both white and Asian students have a higher likelihood of
admission (65% and 60%, respectively) than black and Hispanic students
(57% and 53%). Of those admitted, the racial breakdown is 11% black, 5%
Asian, 2% Hispanic, and 82% white. Not surprisingly, William & Mary Law
School is the most selective of the three, admitting just 21% of all
applicants. Black applicants have the highest chance of admission at
24% compared to 22% for whites and 15% for Asians. Interestingly, the
only large difference is for Hispanics, of whom only 4% are admitted.
The racial breakdown of those admitted the racial breakdown is 8%
black, 9% Asian, 1% Hispanic, and 82% white."
The authors admitted that sometimes ratios are difficult to
conceptualize so they appended a table
Percentage admitted (all applicants are residents):
High School GPA 1051-1150 1151-1200
Black Applicants
SAT Scores 3.3-3.7 86% 86%
3.8-4.2 94% 100%
White Applicants
3.3-3.7 8% 13%
3.8-4.2 37% 51%
.
|
|
|
| User: "tg" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
28 Sep 2005 06:02:41 AM |
|
|
wrote:
I missed your reply hence the delay:
Yes it's stated - see chart 1
"Chart 1 shows the percentage of applicants who were admitted broken
down by
race/ethnic ity without regard to any other characteristic. At UVA,
nearly 65% of all black
applicants were admitted in the Fall of 2003 compared to 36 to 38% of
all other groups. Overall, only 39% of all applicants were admitted. Of
those with race coded, blacks represent about 12% of admissions, Asians
12%, Hispanics 4%, and whites 72%.
To put this in perspective, let's assume a freshman class of 1000.
(Please check my math)
White applicants: 1,946
Black applicants: 185
White admits: 720
Black admits: 120
Black admits at the rate of 37%: 69
'Excess' black admits: 51
Excess black admits as a percentage of the total is 5%
So we have a 5% set-aside for blacks compared to a 100% set-aside for
whites when many of the parents and grandparents of these students were
going to school.
Was that your point?
-tg
NCS is considerably less selective
than UVA, with an overall admission rate of 62%. In addition, the
chance of admissions do not differ as much by race as at UVA, and in
fact both white and Asian students have a higher likelihood of
admission (65% and 60%, respectively) than black and Hispanic students
(57% and 53%). Of those admitted, the racial breakdown is 11% black, 5%
Asian, 2% Hispanic, and 82% white. Not surprisingly, William & Mary Law
School is the most selective of the three, admitting just 21% of all
applicants. Black applicants have the highest chance of admission at
24% compared to 22% for whites and 15% for Asians. Interestingly, the
only large difference is for Hispanics, of whom only 4% are admitted.
The racial breakdown of those admitted the racial breakdown is 8%
black, 9% Asian, 1% Hispanic, and 82% white."
The authors admitted that sometimes ratios are difficult to
conceptualize so they appended a table
Percentage admitted (all applicants are residents):
High School GPA 1051-1150 1151-1200
Black Applicants
SAT Scores 3.3-3.7 86% 86%
3.8-4.2 94% 100%
White Applicants
3.3-3.7 8% 13%
3.8-4.2 37% 51%
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
29 Sep 2005 03:06:29 PM |
|
|
tg wrote:
richasiankid@hotmail.com wrote:
I missed your reply hence the delay:
Yes it's stated - see chart 1
"Chart 1 shows the percentage of applicants who were admitted broken
down by
race/ethnic ity without regard to any other characteristic. At UVA,
nearly 65% of all black
applicants were admitted in the Fall of 2003 compared to 36 to 38% of
all other groups. Overall, only 39% of all applicants were admitted. Of
those with race coded, blacks represent about 12% of admissions, Asians
12%, Hispanics 4%, and whites 72%.
To put this in perspective, let's assume a freshman class of 1000.
(Please check my math)
White applicants: 1,946
Black applicants: 185
White admits: 720
Black admits: 120
Black admits at the rate of 37%: 69
'Excess' black admits: 51
Excess black admits as a percentage of the total is 5%
So we have a 5% set-aside for blacks compared to a 100% set-aside for
whites when many of the parents and grandparents of these students were
going to school.
Was that your point?
-tg
I checked your math and you're correct! But the percentage looks small
only because there is only a fraction of black applicants - blacks are
only 9.5% of the white figure that you cited. So 5% of the "TOTAL" (as
that's the way you chose to express it) is very significant. It's like
saying that more people are killed by cars than tornados, yet this
doesn't mean that cars are more dangerous than tornados. When you drive
a car or approach a tornado, the stats that more people are killed by
cars will not (I hope) tell you that tornados are less dangerous. In
each discrete encounter, the probability matters at that encounter
point and cars are less likely to kill than tornadoes at each
encounter. Cars only kill more often because there are so many of them
around, while tornadoes are rare events.
In this case, if you're a black, 120/185 vs 720/1946 is a significant
difference in percentage. An individual who applies to the university
cares about that admission encounter - about their own discrete
individual probability of being admitted. They may even imagine it
this way - let's say they're comparing themselves one on one -
White applicants: 1000
Black applicants: 1000
White admits: 370
Black admits: 650
Black admits at the rate of 37%: 370
'Excess' black admits: 280
Excess black admits as a percentage of the total is 28% when equal
numbers apply.
Don't forget that SAT scores are not taken into account in the
scenario. The median SAT score of the 720 white admits in this
hypothetical case is 1350, while the median SAT score of the 120 black
admits is 1026, which represents an over 300 point difference which is
quite significant - and depending on how you think about it it can
either be magnified by the above policy, or lead people to think that
it is even more unfair (i.e. in spite of blacks scoring lower, they are
still admitted so much more often)
I think it would be helpful for the paper to provide data on rejected
applicants too which I don't think has been cited (I haven't re-read
it, is it there?)
.
|
|
|
| User: "tg" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
29 Sep 2005 05:26:02 PM |
|
|
Well, at least you refute the stereotype of asian kids being good at
math.
Let's assume that you are an individual white student:
1) Under the existing policy you have a 37/100 chance of being
admitted. (I took the mean of 36 and 38 for our example.)
2) Let's generously give all the 51 set-aside slots to the white
applicants. That would mean that 771 white students would be admitted
out of 1,946. That works out to odds of 39.6/100.
So looking at it from the viewpoint of an *individual* white applicant,
as you suggest, the affirmative action policy decreases your chances of
admission by just about 7 percent. (2.6/37)
If you have a political agenda or you just don't like affirmative
action, that's your business. But I am pointing out your misuse of
numbers. As for the SAT scores, you should look up what 'median' means
and think about comparing the median for 720 people with that for 120.
Are you under the impression that there couldn't be 60 white admits
with scores below 1026?
We will pray that you are not in the hard sciences/engineering.
-tg
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
30 Sep 2005 12:55:01 AM |
|
|
All right! All right!!! Are you saying the following then? Let's say
now we have a policy that mandates that ANY black who applies gets
admitted. In other words, we give a 100% admission rate for blacks. All
you need to get in, if you are black, is to apply. The numbers will
show:
White applicants: 1,946
Black applicants: 185
White admits (at 37%): 720
Black admits (at 100%): 185
Black admits at the rate of 37%*: 68
'Excess' black admits: 117
Excess black admits as a percentage of the total is only ~ slightly
more than 11%. We can easily mandate a 100% black admit policy. The
cost is only 11%!
See your point? What would be the odds ratio comparing blacks to whites
in this case? ;) Again keep in mind SAT scores/GPAs etc are not taken
into account here. "Chart 1 shows the percentage of applicants who were
admitted broken down by race/ethnicity without regard to any other
characteristic. "
As for the SAT part I read elsewhere that in some colleges the
*average* SAT of white rejects is higher than that of the black
matriculants. I'm not saying that it's germane here but since you
mentioned stereotyping......That's why I said it'd be nice to see
reject data as well as matriculant data which I think was not cited in
the paper.
By the way how many blacks do you think scored above the white median
of 1350? Black median is only 1026 as mentioned (based on the 65% admit
rate) and only 60 of blacks scored above 1026. Based on data we know
360 whites scored over *1350* in our hypothetical scenario (I hope I'm
interpreting medians correctly).
FYI I should be honest and I've said elsewhere that I'm quite
indifferent to affirmative action although I have never been able to
bring myself to support it. What's your take on this? In fact after
internet debates I'm starting to have some doubts about it myself as
well meaning as it may appear to be. The data makes it clear that those
who are hurt most are those who are non-black at the bottom of the
class who are rejected (see the data at the end). If you're a high
scorer (GPA/SAT) it makes little difference what race you are and
perhaps so we can be nice and charitable - maybe a certain noblesse
oblige mindset. The more problematic thing is that this social cost is
only relatively contained when there are few blacks applying. If you
have a 50% black and 50% white population applying, then the resultant
cost will be considerably higher to the non-favored group.
*assuming proportionate/"quota"-like representation
.
|
|
|
| User: "tg" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
30 Sep 2005 08:58:53 AM |
|
|
Eventually, I get tired of repeating myself. Whether you like
affirmative action or not is of no interest to me, unless you specify
some goal which you claim it is or isn't meeting.
If you think that your numbers are making a point, you should say what
that point is. All you've managed to demonstrate so far is that if you
treat one group differently from another, you treat one group
differently from another. Well duh.
Statistically, ANY minority group (e.g. people whose names begin with
Q) will show anomalous results if you divide them up amongst a
majority. Look at my post called "education goals" on alt.philosophy
and see if you can follow the numbers.
You also apparently expect that black students with high SAT scores are
going to apply to this crappy school instead of going to a good one
where they can get a scholarship, just so the distribution curve will
look better. I guess you *do* think black people are stupid.
-tg
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
01 Oct 2005 03:02:02 AM |
|
|
In the paper a white applicant with a GPA between 3.3-3.7 with a SAT
between 1051-1150 will only have 8% of getting admitted, while a black
applicant with same qualifications will have a 86% of getting admitted.
The question one needs to focus on is whether blacks and whites should
be held to the same standard in today's world, or should they be held
to a different standard.
In terms of goals, previously in another subthread here I think it was
mentioned that an 'intermediate' goal is college graduation. We also
know that large differences in SAT scores are associated with a large
differential in black-white graduation rate (search the web if you're
not sure about this). Preferences widen SAT differential for obvious
reasons so it's reasonable to hypothesize that this will show up in
grad rates as well, affecting this 'intermediate goal'. Or you can be
more short-sighted and just look at probability of finishing up year 1.
You have illustrated in your own way a point that I never thought about
before - should people from non-favored groups (e.g. whites) avoid
schools with a large population of black applicants (the favored
group)? As you have illustrated the individual cost is proportionate to
the number of blacks applying (and obviously the degree of preference
as well).
Talented blacks will apply to other schools, but the median SAT scores
of whites in elite schools will likely be higher than UVA - you
labelled it a 'crappy school' afterall!! One should expect - if it's
really crappy - that both whites and blacks score low.
I'd like to see data for all applicants (and not just those admitted)
which I've said many times before and unfortunately this has not been
cited. If the only problem is preferential treatment then it is easy to
fix - just change the policy. If there is a disproportionate lack of
talented applicants however (again, see law school admit data in this
paper - there are some hints of a lack of high scoring black
applicants) then this differential will not be easily solved by social
policy. Recall that admitted black median SAT is only 1026 and white
median is 1350 in this instance. An "average" 50th percentile black is
at the very low end if not bottom of the white distribution. (I read
elsewhere blacks are 8-10% percentiles at other schools, at other
places 5%-7%ile of whites in professional schools.)
.
|
|
|
| User: "tg" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
01 Oct 2005 06:22:17 AM |
|
|
wrote:
In the paper a white applicant with a GPA between 3.3-3.7 with a SAT
between 1051-1150 will only have 8% of getting admitted, while a black
applicant with same qualifications will have a 86% of getting admitted.
The question one needs to focus on is whether blacks and whites should
be held to the same standard in today's world, or should they be held
to a different standard.
It isn't the question *I* need to focus on. I need to focus on how
poorly people are educated in science, philosophy, and reason in
general. You obviously didn't read or understand my other thread, and
it may be that you don't have enough reasoning ability, or that you are
a creationist-type person who is incapable of going outside the box of
their social culture.
Like the creationists, (and you may well be one, given the level of
your science education,) you pick out little phrases or bits of
information and throw them around without any comprehension of the
overall system in which they apply.
I will briefly repeat my observation in case there is someone who has
at least .525 of a brain out there.
Take 1000 white students. Pick 100 at random and put a tag on their
folder saying that they are black. Randomly distribute the students
among 4 colleges, where all tests are fair and so on. You will be able
to produce the following headlines at graduation time:
'Black' students do poorly at 3 out of 4 colleges.
'Black' students are not likely to become valedictorians.
'Black' students under-represented in difficult classes.
Until you can understand this, you should just stick to burning
crosses.
-tg
In terms of goals, previously in another subthread here I think it was
mentioned that an 'intermediate' goal is college graduation. We also
know that large differences in SAT scores are associated with a large
differential in black-white graduation rate (search the web if you're
not sure about this). Preferences widen SAT differential for obvious
reasons so it's reasonable to hypothesize that this will show up in
grad rates as well, affecting this 'intermediate goal'. Or you can be
more short-sighted and just look at probability of finishing up year 1.
You have illustrated in your own way a point that I never thought about
before - should people from non-favored groups (e.g. whites) avoid
schools with a large population of black applicants (the favored
group)? As you have illustrated the individual cost is proportionate to
the number of blacks applying (and obviously the degree of preference
as well).
Talented blacks will apply to other schools, but the median SAT scores
of whites in elite schools will likely be higher than UVA - you
labelled it a 'crappy school' afterall!! One should expect - if it's
really crappy - that both whites and blacks score low.
I'd like to see data for all applicants (and not just those admitted)
which I've said many times before and unfortunately this has not been
cited. If the only problem is preferential treatment then it is easy to
fix - just change the policy. If there is a disproportionate lack of
talented applicants however (again, see law school admit data in this
paper - there are some hints of a lack of high scoring black
applicants) then this differential will not be easily solved by social
policy. Recall that admitted black median SAT is only 1026 and white
median is 1350 in this instance. An "average" 50th percentile black is
at the very low end if not bottom of the white distribution. (I read
elsewhere blacks are 8-10% percentiles at other schools, at other
places 5%-7%ile of whites in professional schools.)
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
01 Oct 2005 02:39:37 PM |
|
|
Very patient sigh........
I've been trying to discuss this rationally with this thread but it
seems that you have a lot of emotive baggage in your posts - starting
from "*at least you refute* the Asian stereotype" to "we'll pray that
you're not in the hard sciences" to "I guess you *do* think black
people are stupid" etc, and now you cite "creationists" and then
"burning crosses"?? What?? I must admit it's hard to follow!! (hey,
by the way, is that how philosophers think these days? A friend said
that's why many philosophers are poorly paid rightly or wrongly;
society has little use of them!! I give them benefit of the doubt but
still another guy I met said some philosophical departments are
undergoing 'self-liquidation' whatever that means. Hmmmmmmmm.)
This will therefore be my *last* post. (You can stand up and take a
bow and cry 'hoooray! Kick him outta here!! Stifle his unpleasant
facts! I don't like what I see so "*I* need to focus elsewhere"!)
Hehehehehe....
Let's recap. From the paper, and with numbers you provided assuming a
freshman class of 1000:
At UVA, white median is 1350 and a whopping 360 whites scored above and
below this threshold
Black median is 1026 and only 60 blacks scored above and below this
threshold
If you pick a black at random, 50% of the time he will score below
1026. This is exceedingly low in the white distribution -- recall that
very few whites who score between 1050-1150 get admitted in the first
place. Half of black students are at the very very bottom. This
contributes significantly to 'stereotyping' on college campus at UVA I
suspect. But this 'stereotyping' is based on some element truth as we
just saw - they really have low scores. The distributions of blacks and
whites in college are not identical and in fact differ profoundly and
astonishingly. Black median is lower than the white median by more than
300 SAT points. Are you suggesting that this is not a huge difference?
We've also seen that blacks and whites are held to very different
criteria during the admission process. The odds ratio for admission for
blacks compared to whites (with SAT/LSAT scores, gpa, and residence
controlled for) at UVA is 106 to 1. The corresponding odds ratio of
admit at William &Mary law school is 267 to 1. In other words, at UVA
the odds of a black student being admitted is more than 100 times the
odds of admission of a white student with the same qualifications. The
odds of admitting a black applicant at William & Mary is more than 250
times the odds of admitting an equally-qualified white applicant.
That's all in the paper - I didn't make this up!
In the stats you provided you implied perhaps that even if we admit ALL
blacks, i.e. 100% blacks, i.e. *any black who applies gets in* without
regard to merit, GPAs, SATs etc - the cost to the total would only be
slightly over 11% (backtrack a few posts if you don't remember this),
since they are only a small percentage of applicants. I think most
people will have concerns about this even if overall cost is 11% -
what, blacks get 100% admit rate regardless of merit?? Anyone who
applies gets in??? In that case, what do you think people will say
when they see black students on the campus?
Now, when the percentage of blacks who apply increases, so does the
corresponding social cost. Should people stay away from colleges with a
huge black applicant population as they will be treated as second-class
applicants? Well? That's one interpretation of the data.
Of course, the difference between the above scenario and in real life
is only a matter of degree. It does not differ in fundamentals. As we
have seen, standards have already been thrown out the window with an
over 300 median SAT point difference in the freshman class.
The pertinent question therefore is whether people should be held to
the same standard, or whether people should be held to such a
drastically different standard in the UVA admit process. What is your
take on this? Holding people to different standards contributes (albeit
not totally) to huge median SAT (and likely LSAT) differentials between
favored (blacks) and nonfavored groups (all others), with visible
ramifications in campus life and dropout rates. The more important
indictment of such a practice, however, is that it deeply conflicts
with principles of equal treatment, equality and fairness that we
cherish.
And with this I close our discussion. Nice chatting with you,
sincerely,
---
addendum: Your little thought experiment - were you suggesting that
some may benefit from majority or all black schools or universities?
Howard University, the "black Harvard" as some call it, cites a mean
SAT of only 1019. A quarter of students score under 820. Was it you who
said that UVA was a "crappy" school? ;)
.
|
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
01 Oct 2005 02:03:50 PM |
|
|
Very patient sigh........
I've been trying to discuss this rationally with this thread but it
seems that you have a lot of emotive baggage in your posts - starting
from "*at least you refute* the Asian stereotype" to "we'll pray that
you're not in the hard sciences" to "I guess you *do* think black
people are stupid" etc, and now you cite "creationists" and then
"burning crosses"?? What?? I must admit it's hard to follow!! (hey,
by the way, is that how philosophers think these days? A friend said
that's why many philosophers are poorly paid rightly or wrongly;
society has little use of them!! I give them benefit of the doubt but
still another guy I met said some philosophical departments are
undergoing 'self-liquidation' whatever that means. Hmmmmmmmm.)
This will therefore be my *last* post. (You can stand up and take a
bow and cry 'hoooray! Kick him outta here!! Stifle his unpleasant
facts! I don't like what I see so "*I* need to focus elsewhere"!)
Hehehehehe....
Let's recap. From the paper, and with numbers you provided assuming a
freshman class of 1000:
At UVA, white median is 1350 and a whopping 360 whites scored above and
below this threshold
Black median is 1026 and only 60 blacks scored above and below this
threshold
If you pick a black at random, 50% of the time he will score below
1026. This is exceedingly low in the white distribution -- recall that
very few whites who score between 1050-1150 get admitted in the first
place. Half of black students are at the very very bottom. This
contributes significantly to 'stereotyping' on college campus at UVA I
suspect. But this 'stereotyping' is based on some element truth as we
just saw - they really have low scores. The distributions of blacks and
whites in college are not identical and in fact differ profoundly and
astonishingly. Black median is lower than the white median by more than
300 SAT points. Are you suggesting that this is not a huge difference?
We've also seen that blacks and whites are held to very different
criteria during the admission process. The odds ratio for admission for
blacks compared to whites (with SAT/LSAT scores, gpa, and residence
controlled for) at UVA is 106 to 1. The corresponding odds ratio of
admit at William &Mary law school is 267 to 1. In other words, at UVA
the odds of a black student being admitted is more than 100 times the
odds of admission of a white student with the same qualifications. The
odds of admitting a black applicant at William & Mary is more than 250
times the odds of admitting an equally-qualified white applicant.
That's all in the paper - I didn't make this up!
In the stats you provided you implied perhaps that even if we admit ALL
blacks, i.e. 100% blacks, i.e. *any black who applies gets in* without
regard to merit, GPAs, SATs etc - the cost to the total would only be
slightly over 11% (backtrack a few posts if you don't remember this),
since they are only a small percentage of applicants. I think most
people will have concerns about this even if overall cost is 11% -
what, blacks get 100% admit rate regardless of merit?? Anyone who
applies gets in??? In that case, what do you think people will say
when they see black students on the campus?
Now, when the percentage of blacks who apply increases, so does the
corresponding social cost. Should people stay away from colleges with a
huge black applicant population as they will be treated as second-class
applicants? Well? That's one interpretation of the data.
Of course, the difference between the above scenario and in real life
is only a matter of degree. It does not differ in fundamentals. As we
have seen, standards have already been thrown out the window with an
over 300 median SAT point difference in the freshman class.
The pertinent question therefore is whether people should be held to
the same standard, or whether people should be held to such a
drastically different standard in the UVA admit process. I'm inclined
to say no, in part because this contributes (albeit not totally) to
huge median SAT (and likely LSAT) differentials between favored
(blacks) and nonfavored groups (all others), with visible ramifications
in campus life and dropout rates. The more important indictment,
however, is such a practice deeply conflicts with principles of
equality and fairness that we cherish.
And with this I close our discussion. Nice chatting with you,
sincerely,
---
addendum: Your little thought experiment - were you suggesting that
some may benefit from majority or all black schools or universities?
Howard University, the "black Harvard" as some call it, cites a mean
SAT of only 1019. A quarter of students score under 820. Was it you who
said that UVA was a "crappy" school? ;)
.
|
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
01 Oct 2005 02:00:06 PM |
|
|
Very patient sigh........
I've been trying to discuss this rationally with this thread but it
seems that you have a lot of emotive baggage in your posts - starting
from "*at least you refute* the Asian stereotype" to "we'll pray that
you're not in the hard sciences" to "I guess you *do* think black
people are stupid" etc, and now you cite "creationists" and then
"burning crosses"?? What?? I must admit it's hard to follow!! (hey,
by the way, is that how philosophers think these days? A friend said
that's why many philosophers are poorly paid rightly or wrongly;
society has little use of them!! I give them benefit of the doubt but
still another guy I met said some philosophical departments are
undergoing 'self-liquidation' whatever that means. Hmmmmmmmm.)
This will therefore be my *last* post. (You can stand up and take a
bow and cry 'hoooray! Kick him outta here!! Stifle his unpleasant
facts! I don't like what I see so "*I* need to focus elsewhere"!)
Hehehehehe....
Let's recap. From the paper, and with numbers you provided assuming a
freshman class of 1000:
At UVA, white median is 1350 and a whopping 360 whites scored above and
below this threshold
Black median is 1026 and only 60 blacks scored above and below this
threshold
If you pick a black at random, 50% of the time he will score below
1026. This is exceedingly low in the white distribution -- recall that
very few whites who score between 1050-1150 get admitted in the first
place. Half of black students are at the very very bottom. This
contributes significantly to 'stereotyping' on college campus at UVA I
suspect. But this 'stereotyping' is based on some element truth as we
just saw - they really have low scores. The distributions of blacks and
whites in college are not identical and in fact differ profoundly and
astonishingly. Black median is lower than the white median by more than
300 SAT points. Are you suggesting that this is not a huge difference?
We've also seen that blacks and whites are held to very different
criteria during the admission process. The odds ratio for admission for
blacks compared to whites (with SAT/LSAT scores, gpa, and residence
controlled for) at UVA is 106 to 1. The corresponding odds ratio of
admit at William &Mary law school is 267 to 1. In other words, at UVA
the odds of a black student being admitted is more than 100 times the
odds of admission of a white student with the same qualifications. The
odds of admitting a black applicant at William & Mary is more than 250
times the odds of admitting an equally-qualified white applicant.
That's all in the paper - I didn't make this up!
In the stats you provided you implied perhaps that even if we admit ALL
blacks, i.e. 100% blacks, i.e. *any black who applies gets in* without
regard to merit, GPAs, SATs etc - the cost to the total would only be
slightly over 11% (backtrack a few posts if you don't remember this),
since they are only a small percentage of applicants. I think most
people will have concerns about this even if overall cost is 11% -
what, blacks get 100% admit rate regardless of merit?? Anyone who
applies gets in??? In that case, what do you think people will say
when they see black students on the campus?
Now, when the percentage of blacks who apply increases, so does the
corresponding social cost. Should people stay away from colleges with a
huge black applicant population as they will be treated as second-class
applicants? Well? That's one interpretation of the data.
Of course, the difference between the above scenario and in real life
is only a matter of degree. It does not differ in fundamentals. As we
have seen, standards have already been thrown out the window with an
over 300 median SAT point difference in the freshman class.
The pertinent question therefore is whether people should be held to
the same standard, or whether people should be held to such a
drastically different standard in the UVA admit process. I'm inclined
to say no, in part because this contributes (albeit not totally) to
huge median SAT (and likely LSAT) differentials between favored
(blacks) and nonfavored groups (all others), with visible ramifications
in campus life and dropout rates. The more important indictment,
however, is such a practice deeply conflicts with principles of
equality and fairness that we cherish.
Nice chatting with you, sincerely,
---
addendum: Your little thought experiment - were you suggesting that
some may benefit from majority or all black schools or universities?
Howard University, the "black Harvard" as some call it, cites a mean
SAT of only 1019. A quarter of students score under 820. Was it you who
said that UVA was a "crappy" school? ;)
tg wrote:
richasiankid@hotmail.com wrote:
In the paper a white applicant with a GPA between 3.3-3.7 with a SAT
between 1051-1150 will only have 8% of getting admitted, while a black
applicant with same qualifications will have a 86% of getting admitted.
The question one needs to focus on is whether blacks and whites should
be held to the same standard in today's world, or should they be held
to a different standard.
It isn't the question *I* need to focus on. I need to focus on how
poorly people are educated in science, philosophy, and reason in
general. You obviously didn't read or understand my other thread, and
it may be that you don't have enough reasoning ability, or that you are
a creationist-type person who is incapable of going outside the box of
their social culture.
Like the creationists, (and you may well be one, given the level of
your science education,) you pick out little phrases or bits of
information and throw them around without any comprehension of the
overall system in which they apply.
I will briefly repeat my observation in case there is someone who has
at least .525 of a brain out there.
Take 1000 white students. Pick 100 at random and put a tag on their
folder saying that they are black. Randomly distribute the students
among 4 colleges, where all tests are fair and so on. You will be able
to produce the following headlines at graduation time:
'Black' students do poorly at 3 out of 4 colleges.
'Black' students are not likely to become valedictorians.
'Black' students under-represented in difficult classes.
Until you can understand this, you should just stick to burning
crosses.
-tg
In terms of goals, previously in another subthread here I think it was
mentioned that an 'intermediate' goal is college graduation. We also
know that large differences in SAT scores are associated with a large
differential in black-white graduation rate (search the web if you're
not sure about this). Preferences widen SAT differential for obvious
reasons so it's reasonable to hypothesize that this will show up in
grad rates as well, affecting this 'intermediate goal'. Or you can be
more short-sighted and just look at probability of finishing up year 1.
You have illustrated in your own way a point that I never thought about
before - should people from non-favored groups (e.g. whites) avoid
schools with a large population of black applicants (the favored
group)? As you have illustrated the individual cost is proportionate to
the number of blacks applying (and obviously the degree of preference
as well).
Talented blacks will apply to other schools, but the median SAT scores
of whites in elite schools will likely be higher than UVA - you
labelled it a 'crappy school' afterall!! One should expect - if it's
really crappy - that both whites and blacks score low.
I'd like to see data for all applicants (and not just those admitted)
which I've said many times before and unfortunately this has not been
cited. If the only problem is preferential treatment then it is easy to
fix - just change the policy. If there is a disproportionate lack of
talented applicants however (again, see law school admit data in this
paper - there are some hints of a lack of high scoring black
applicants) then this differential will not be easily solved by social
policy. Recall that admitted black median SAT is only 1026 and white
median is 1350 in this instance. An "average" 50th percentile black is
at the very low end if not bottom of the white distribution. (I read
elsewhere blacks are 8-10% percentiles at other schools, at other
places 5%-7%ile of whites in professional schools.)
.
|
|
|
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
01 Oct 2005 01:14:22 AM |
|
|
In the paper a white applicant with a GPA between 3.3-3.7 with a SAT
between 1051-1150 will only have 8% of getting admitted, while a black
applicant with same qualifications will have a 86% of getting admitted.
The question one needs to focus on is whether blacks and whites should
be held to the same standard in today's world, or should they be held
to a different standard.
In terms of goals, previously in another subthread here I think it was
mentioned that an 'intermediate' goal is college graduation. We also
know that large differences in SAT scores are associated with a large
differential in black-white graduation rate (search the web if you're
not sure about this). Preferences widen SAT differential for obvious
reasons so it's reasonable to hypothesize that this will show up in
grad rates as well, affecting this 'intermediate goal'. Or you can be
more short-sighted and just look at probability of finishing up year 1.
You have illustrated in your own way a point that I never thought about
before - should people from non-favored groups (e.g. whites) avoid
schools with a large population of black applicants (the favored
group)? As you have illustrated the individual cost is proportionate to
the number of blacks applying (and obviously the degree of preference
as well).
Talented blacks will apply to other schools, but the median SAT scores
of whites in elite schools will likely be higher than UVA - you
labelled it a 'crappy school' afterall!! One should expect - if it's
really crappy - that both whites and blacks score low.
I'd like to see data for all applicants (and not just those admitted)
which I've said many times before and unfortunately this has not been
cited. If the only problem is preferential treatment then it is easy to
fix - just change the policy. If there is a disproportionate lack of
talented applicants however (again, see law school admit data in this
paper - there are some hints of a lack of high scoring black
applicants) then this differential will not be easily solved by social
policy. Regardless students will see through their eyes their own ways
when there is a disproportionate black drop out rate - recall that
admitted black median SAT is only 1026 and white median is 1350 in this
instance. A 50th percentile black is at the very low end if not bottom
of the white distribution. (I read elsewhere blacks are 8-10%
percentiles at other schools, at other places 5%-7%ile of whites in
professional schools. )
.
|
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
01 Oct 2005 01:09:30 AM |
|
|
In the paper a white applicant with a GPA between 3.3-3.7 with a SAT
between 1051-1150 will only have 8% of getting admitted, while a black
applicant with same qualifications will have a 86% of getting admitted.
The question one needs to focus on is whether blacks and whites should
be held to the same standard in today's world, or should they be held
to a different standard.
In terms of goals, previously in another subthread here I think it was
mentioned that an 'intermediate' goal is college graduation. We also
know that large differences in SAT scores are associated with a large
differential in black-white graduation rate (search the web if you're
not sure about this). Preferences widen SAT differential for obvious
reasons so it's reasonable to hypothesize that this will show up in
grad rates as well, affecting this 'intermediate goal'. Or you can be
more short-sighted and just look at probability of finishing up year 1.
You have illustrated in your own way a point that I never thought about
before - should people from non-favored groups (e.g. whites) avoid
schools with a large population of black applicants (the favored
group)? As you have illustrated the individual cost is proportionate to
the number of blacks applying (and obviously the degree of preference
as well).
Talented blacks will apply to other schools, but the median SAT scores
of whites in elite schools will likely be higher than UVA - you
labelled it a 'crappy school' afterall!! One should expect - if it's
really crappy - that both whites and blacks score low.
I'd like to see data for all applicants (and not just those admitted)
which I've said many times before and unfortunately this has not been
cited. If the only problem is preferential treatment then it is easy to
fix - just change the policy. If there is a disproportionate lack of
talented applicants however (again, see law school admit data in this
paper - there are some hints of a lack of high scoring black
applicants) then this differential will not be easily solved by social
policy. Regardless students will see through their eyes their own ways
when there is a disproportionate black drop out rate - recall that
admitted black median SAT is only 1026 and white median is 1350 in this
instance. A 50th percentile black is at the very low end if not bottom
of the white distribution. (I read elsewhere blacks are 8-10%
percentiles at other schools, at other places 5%-7%ile of whites in
professional schools. We don't have enough data here to make inference
on percentiles in this particular school.)
.
|
|
|
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
29 Sep 2005 08:16:12 PM |
|
|
All right, all right, all right!!!! Are you saying the following then?
Let's say now we mandate that ANY black who applies gets admitted. In
other words, we give a 100% admission rate for blacks. All you need to
get in, if you are black, is to apply. What do we now have?
White applicants: 1,946
Black applicants: 185
White admits (at 37%): 720
Black admits (at 100%): 185
Black admits at the rate of 37%: 68
'Excess' black admits: 117
Excess black admits as a percentage of the total is only ~ slightly
more than 11%!! We can easily mandate a 100% black admit policy. The
cost is only 11%!!
See your point? What would be the odds ratio in that case? ;)
As for the SAT part I read elsewhere that in some colleges the
*average* SAT of white rejects is higher than that of the black
matriculants. I'm not saying that it's germane here but since you
mentioned stereotyping....That's why I said it'd be nice to see reject
data as well as matriculant data which I think was not cited in the
paper.
By the way how many blacks do you think scored above the white median
of 1350? Black median is only 1026 as mentioned and only 60 of blacks
scored above this point. Based on data we know 360 whites scored over
*1350* in our hypothetical scenario (I hope I'm interpreting medians
correctly)
FYI I should be honest and I've said elsewhere that I'm quite callous
about affirmative action although I cannot bring myself to support it.
In fact after internet debates I'm starting to have doubts about it
myself. The data makes it clear that those who are hurt most are those
who are non-black at the bottom of the class who are rejected (see the
data at the end). If you're a high scorer (GPA/SAT) it makes no
difference what race you are and you still get in any way. I suspect
being nice and charitable has to do with a certain noblesse oblige
mindset. The more problematic thing is that this social cost is only
contained when there are relatively few blacks applying. If you have a
50% black and 50% white population applying then the cost will be
considerably higher to the non-favored group.
tg wrote:
Well, at least you refute the stereotype of asian kids being good at
math.
Let's assume that you are an individual white student:
1) Under the existing policy you have a 37/100 chance of being
admitted. (I took the mean of 36 and 38 for our example.)
2) Let's generously give all the 51 set-aside slots to the white
applicants. That would mean that 771 white students would be admitted
out of 1,946. That works out to odds of 39.6/100.
So looking at it from the viewpoint of an *individual* white applicant,
as you suggest, the affirmative action policy decreases your chances of
admission by just about 7 percent. (2.6/37)
If you have a political agenda or you just don't like affirmative
action, that's your business. But I am pointing out your misuse of
numbers. As for the SAT scores, you should look up what 'median' means
and think about comparing the median for 720 people with that for 120.
Are you under the impression that there couldn't be 60 white admits
with scores below 1026?
We will pray that you are not in the hard sciences/engineering.
-tg
.
|
|
|
|
|
|
|
|
|
|
|
| User: "Edgar Svendsen" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
22 Sep 2005 08:20:44 AM |
|
|
<richasiankid@hotmail.com> wrote in message
news:1127341358.848521.111510@z14g2000cwz.googlegroups.com...
http://www.ceousa.org/pdfs/VAS%20Report.pdf (you need Adobe Reader to
view it)
"The odds ratio for blacks compared to whites at NCS is 13 to 1, but at
UVA it is 106 to 1 and at William &Mary 267 to 1. In other words, at
UVA the odds of a black student being admitted is more than 100 times
the odds of admission of a white student with the same qualifications.
The odds of admitting a black applicant at William & Mary is more than
250 times the odds of admitting an equally-qualified white applicant.
Does this mean that very few blacks apply to William & Mary and a great many
whites do apply?
How did they determine that the odds apply to equally-qualified applicants.
Does this mean that, if tested, the successful black applicants would test
as exactly equally-qualified as the successful white applicants?
Ed
The odds ratios for Asians at all three schools are less than one,
meaning that Asians are less likely to be admitted than
equally-qualified whites (the odds ratio for Asians at UVA is not
statistically significant). The odds ratios for Hispanics are 2.8 and
1.9 at UVA and NCS, respectively, but less than one at W&M. This means
that Hispanics are somewhat more likely than whites to be admitted at
UVA and NCS but less likely than whites to be admitted at W&M Law (but
this is not statistically significant)."
University of Virginia (UVA) undergraduates, North Carolina State (NCS)
undergraduates, and William and Mary Law (W&M Law) School.
.
|
|
|
| User: "tg" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
22 Sep 2005 12:07:15 PM |
|
|
Edgar Svendsen wrote:
<richasiankid@hotmail.com> wrote in message
news:1127341358.848521.111510@z14g2000cwz.googlegroups.com...
http://www.ceousa.org/pdfs/VAS%20Report.pdf (you need Adobe Reader to
view it)
"The odds ratio for blacks compared to whites at NCS is 13 to 1, but at
UVA it is 106 to 1 and at William &Mary 267 to 1. In other words, at
UVA the odds of a black student being admitted is more than 100 times
the odds of admission of a white student with the same qualifications.
The odds of admitting a black applicant at William & Mary is more than
250 times the odds of admitting an equally-qualified white applicant.
Does this mean that very few blacks apply to William & Mary and a great many
whites do apply?
How did they determine that the odds apply to equally-qualified applicants.
Does this mean that, if tested, the successful black applicants would test
as exactly equally-qualified as the successful white applicants?
Ed
Faint chance of a rational discussion on this, but:
1) There are no tests.
2) You can't isolate the variables.
People who are serious about these issues have to identify a *goal*
first, and then test to see if the goal is achieved, given their
theoretical predictions.
The SAT is an example of this, although it is widely misinterpreted.
The test predicts whether a student will finish the first year of
college (and that's all). This is valuable to colleges because
first-year success means the student is likely to come back and give
them more money, and they don't have to recruit second-year transfers,
which can be difficult and costly.
In making racial comparisons, isolating variables is extremely
difficult. There is no 'control group'; having identical SAT scores for
a black and white student is meaningless, since there are so many other
factors at work which influence the outcome, like peer and family
social networks.
-tg
.
|
|
|
| User: "Edgar Svendsen" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
22 Sep 2005 05:03:53 PM |
|
|
"tg" <tgdenning@earthlink.net> wrote in message
news:1127408835.355207.59100@o13g2000cwo.googlegroups.com...
Edgar Svendsen wrote:
<richasiankid@hotmail.com> wrote in message
news:1127341358.848521.111510@z14g2000cwz.googlegroups.com...
http://www.ceousa.org/pdfs/VAS%20Report.pdf (you need Adobe Reader to
view it)
"The odds ratio for blacks compared to whites at NCS is 13 to 1, but at
UVA it is 106 to 1 and at William &Mary 267 to 1. In other words, at
UVA the odds of a black student being admitted is more than 100 times
the odds of admission of a white student with the same qualifications.
The odds of admitting a black applicant at William & Mary is more than
250 times the odds of admitting an equally-qualified white applicant.
Does this mean that very few blacks apply to William & Mary and a great
many
whites do apply?
How did they determine that the odds apply to equally-qualified
applicants.
Does this mean that, if tested, the successful black applicants would
test
as exactly equally-qualified as the successful white applicants?
Ed
Faint chance of a rational discussion on this, but:
1) There are no tests.
2) You can't isolate the variables.
People who are serious about these issues have to identify a *goal*
first, and then test to see if the goal is achieved, given their
theoretical predictions.
The SAT is an example of this, although it is widely misinterpreted.
The test predicts whether a student will finish the first year of
college (and that's all). This is valuable to colleges because
first-year success means the student is likely to come back and give
them more money, and they don't have to recruit second-year transfers,
which can be difficult and costly.
In making racial comparisons, isolating variables is extremely
difficult. There is no 'control group'; having identical SAT scores for
a black and white student is meaningless, since there are so many other
factors at work which influence the outcome, like peer and family
social networks.
True, of course. The more information you have the better judgement you can
make, even if the information is incomplete. It would be good to know the
total number of white and black, and other applicants at each of these
schools. Clearly the schools are trying to favor groups seen as
disadvantaged; to judge the success of these efforts would require more
data than richasiankid gave us, like graduation rates too
Ed
-tg
.
|
|
|
| User: "tg" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
22 Sep 2005 06:06:02 PM |
|
|
Edgar Svendsen wrote:
"tg" <tgdenning@earthlink.net> wrote in message
news:1127408835.355207.59100@o13g2000cwo.googlegroups.com...
Edgar Svendsen wrote:
<richasiankid@hotmail.com> wrote in message
news:1127341358.848521.111510@z14g2000cwz.googlegroups.com...
http://www.ceousa.org/pdfs/VAS%20Report.pdf (you need Adobe Reader to
view it)
"The odds ratio for blacks compared to whites at NCS is 13 to 1, but at
UVA it is 106 to 1 and at William &Mary 267 to 1. In other words, at
UVA the odds of a black student being admitted is more than 100 times
the odds of admission of a white student with the same qualifications.
The odds of admitting a black applicant at William & Mary is more than
250 times the odds of admitting an equally-qualified white applicant.
Does this mean that very few blacks apply to William & Mary and a great
many
whites do apply?
How did they determine that the odds apply to equally-qualified
applicants.
Does this mean that, if tested, the successful black applicants would
test
as exactly equally-qualified as the successful white applicants?
Ed
Faint chance of a rational discussion on this, but:
1) There are no tests.
2) You can't isolate the variables.
People who are serious about these issues have to identify a *goal*
first, and then test to see if the goal is achieved, given their
theoretical predictions.
The SAT is an example of this, although it is widely misinterpreted.
The test predicts whether a student will finish the first year of
college (and that's all). This is valuable to colleges because
first-year success means the student is likely to come back and give
them more money, and they don't have to recruit second-year transfers,
which can be difficult and costly.
In making racial comparisons, isolating variables is extremely
difficult. There is no 'control group'; having identical SAT scores for
a black and white student is meaningless, since there are so many other
factors at work which influence the outcome, like peer and family
social networks.
True, of course. The more information you have the better judgement you can
make, even if the information is incomplete. It would be good to know the
total number of white and black, and other applicants at each of these
schools. Clearly the schools are trying to favor groups seen as
disadvantaged; to judge the success of these efforts would require more
data than richasiankid gave us, like graduation rates too
Ed
But you haven't defined a *goal* or desired outcome. What is "success
of these efforts"? Is the goal to have a
general-population-proportionate number of black students graduate?
-tg
-tg
.
|
|
|
| User: "Edgar Svendsen" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
23 Sep 2005 07:21:00 PM |
|
|
"tg" <tgdenning@earthlink.net> wrote in message
news:1127430361.974501.119830@g14g2000cwa.googlegroups.com...
Edgar Svendsen wrote:
"tg" <tgdenning@earthlink.net> wrote in message
news:1127408835.355207.59100@o13g2000cwo.googlegroups.com...
Edgar Svendsen wrote:
<richasiankid@hotmail.com> wrote in message
news:1127341358.848521.111510@z14g2000cwz.googlegroups.com...
http://www.ceousa.org/pdfs/VAS%20Report.pdf (you need Adobe Reader
to
view it)
"The odds ratio for blacks compared to whites at NCS is 13 to 1, but
at
UVA it is 106 to 1 and at William &Mary 267 to 1. In other words, at
UVA the odds of a black student being admitted is more than 100
times
the odds of admission of a white student with the same
qualifications.
The odds of admitting a black applicant at William & Mary is more
than
250 times the odds of admitting an equally-qualified white
applicant.
Does this mean that very few blacks apply to William & Mary and a
great
many
whites do apply?
How did they determine that the odds apply to equally-qualified
applicants.
Does this mean that, if tested, the successful black applicants would
test
as exactly equally-qualified as the successful white applicants?
Ed
Faint chance of a rational discussion on this, but:
1) There are no tests.
2) You can't isolate the variables.
People who are serious about these issues have to identify a *goal*
first, and then test to see if the goal is achieved, given their
theoretical predictions.
The SAT is an example of this, although it is widely misinterpreted.
The test predicts whether a student will finish the first year of
college (and that's all). This is valuable to colleges because
first-year success means the student is likely to come back and give
them more money, and they don't have to recruit second-year transfers,
which can be difficult and costly.
In making racial comparisons, isolating variables is extremely
difficult. There is no 'control group'; having identical SAT scores for
a black and white student is meaningless, since there are so many other
factors at work which influence the outcome, like peer and family
social networks.
True, of course. The more information you have the better judgement you
can
make, even if the information is incomplete. It would be good to know
the
total number of white and black, and other applicants at each of these
schools. Clearly the schools are trying to favor groups seen as
disadvantaged; to judge the success of these efforts would require more
data than richasiankid gave us, like graduation rates too
Ed
But you haven't defined a *goal* or desired outcome. What is "success
of these efforts"? Is the goal to have a
general-population-proportionate number of black students graduate?
No, the goal is to reduce the disparity in income between the general white
population and the minority populations. One tool to accomplish this goal
is to have as many minority students as is economically and politcally
feasible graduate from college and get higher paying jobs.
Ed
-tg
-tg
.
|
|
|
| User: "tg" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
24 Sep 2005 07:11:05 AM |
|
|
Edgar Svendsen wrote:
"tg" <tgdenning@earthlink.net> wrote in message
news:1127430361.974501.119830@g14g2000cwa.googlegroups.com...
Edgar Svendsen wrote:
"tg" <tgdenning@earthlink.net> wrote in message
news:1127408835.355207.59100@o13g2000cwo.googlegroups.com...
Edgar Svendsen wrote:
<richasiankid@hotmail.com> wrote in message
news:1127341358.848521.111510@z14g2000cwz.googlegroups.com...
http://www.ceousa.org/pdfs/VAS%20Report.pdf (you need Adobe Reader
to
view it)
"The odds ratio for blacks compared to whites at NCS is 13 to 1, but
at
UVA it is 106 to 1 and at William &Mary 267 to 1. In other words, at
UVA the odds of a black student being admitted is more than 100
times
the odds of admission of a white student with the same
qualifications.
The odds of admitting a black applicant at William & Mary is more
than
250 times the odds of admitting an equally-qualified white
applicant.
Does this mean that very few blacks apply to William & Mary and a
great
many
whites do apply?
How did they determine that the odds apply to equally-qualified
applicants.
Does this mean that, if tested, the successful black applicants would
test
as exactly equally-qualified as the successful white applicants?
Ed
Faint chance of a rational discussion on this, but:
1) There are no tests.
2) You can't isolate the variables.
People who are serious about these issues have to identify a *goal*
first, and then test to see if the goal is achieved, given their
theoretical predictions.
The SAT is an example of this, although it is widely misinterpreted.
The test predicts whether a student will finish the first year of
college (and that's all). This is valuable to colleges because
first-year success means the student is likely to come back and give
them more money, and they don't have to recruit second-year transfers,
which can be difficult and costly.
In making racial comparisons, isolating variables is extremely
difficult. There is no 'control group'; having identical SAT scores for
a black and white student is meaningless, since there are so many other
factors at work which influence the outcome, like peer and family
social networks.
True, of course. The more information you have the better judgement you
can
make, even if the information is incomplete. It would be good to know
the
total number of white and black, and other applicants at each of these
schools. Clearly the schools are trying to favor groups seen as
disadvantaged; to judge the success of these efforts would require more
data than richasiankid gave us, like graduation rates too
Ed
But you haven't defined a *goal* or desired outcome. What is "success
of these efforts"? Is the goal to have a
general-population-proportionate number of black students graduate?
No, the goal is to reduce the disparity in income between the general white
population and the minority populations. One tool to accomplish this goal
is to have as many minority students as is economically and politcally
feasible graduate from college and get higher paying jobs.
Ed
Ok, but the graduation rate is the intermediate goal. I started another
thread on this--- "Education Goals"--- which may explain my point. I
suspect that if we held all else constant, we would see your ultimate
goal to be met through affirmative action, but we would not see the
intermediate goal.
-tg
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
24 Sep 2005 02:42:36 PM |
|
|
I read that even students who have completed some college do better
economically than those who only finish high school, supporting your
point that the ultimate goal is achieved while the intermediate goal
isn't necessarily (ie blacks dropping out more often because they're
not as academically capable as perhaps predicted by the SAT, but at
least they'd have it on paper saying they've gone to some college).
But as you said elsewhere finishing 1st year is predicted by the SAT,
and that students are more likely to come back 2nd year etc. Will that
make it more likely, however infinitesimally, that they go on 3rd year
and 4th? It'd be great to do a prospective study on this.
.
|
|
|
| User: "tg" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
25 Sep 2005 05:45:55 AM |
|
|
First, let me point out that to get out my message---"it isn't that
simple"---I simplified a bit. SAT tells us that one will do ok
academically, but some people drop out for other reasons. That said,
the highest rate of attrition *is* between first and second year. Of
course, things vary from institution to institution.
In terms of Ed's overall goal, indeed, affirmative action works even
under editor@netpath' s negative interpretation, since it *does* create
educated black women who can move into the middle class. Not to
mention Colin Powell, and all the others who are not as visible due to
the statistical effects discussed in the other thread.
-tg
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
25 Sep 2005 07:40:18 AM |
|
|
TG wrote:
In terms of Ed's overall goal, indeed, affirmative action works even
under editor@netpath' s negative interpretation, since it *does* create
educated black women who can move into the middle class.
No it doesn't "work" - as there now is a gross imbalance in America
between educated middle-class black women and educated middle-class
black men for them to marry, an issue that even the very-liberal
Washington Post has run big stories on.
Face it, the fastest, most-cost-effective route to the middle class
- for any man or woman, of any race - is COMMUNITY COLLEGE. Nil costs.
Graduate in two years to a high-paying career as something like an
electrician, plumber, nurse, gunsmith, welder - jobs that can't be
"offshored" to India, unlike your middle-management twin brother who
now has that big student-loan debt from prestigious four-year college.
No $4 to park! No $6 admission!
http://stores.ebay.com/INTERNET-GUN-SHOW
.
|
|
|
| User: "tg" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
25 Sep 2005 09:01:15 AM |
|
|
wrote:
TG wrote:
In terms of Ed's overall goal, indeed, affirmative action works even
under editor@netpath' s negative interpretation, since it *does* create
educated black women who can move into the middle class.
No it doesn't "work" - as there now is a gross imbalance in America
between educated middle-class black women and educated middle-class
black men for them to marry,
Ed's goal was to have more minority people in the middle class. Of
course, if you are a racist, you believe that sex and marriage between
middle class white men and middle class black women is sinful or some
such fundamentalist nonsense. And I'm sitting here trying to educate
you about statistics. Sheesh.
-tg
an issue that even the very-liberal
Washington Post has run big stories on.
Face it, the fastest, most-cost-effective route to the middle class
- for any man or woman, of any race - is COMMUNITY COLLEGE. Nil costs.
Graduate in two years to a high-paying career as something like an
electrician, plumber, nurse, gunsmith, welder - jobs that can't be
"offshored" to India, unlike your middle-management twin brother who
now has that big student-loan debt from prestigious four-year college.
No $4 to park! No $6 admission!
http://stores.ebay.com/INTERNET-GUN-SHOW
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
25 Sep 2005 01:22:14 PM |
|
|
TG wrote:
In terms of Ed's overall goal, indeed, affirmative action works even
under editor@netpath' s negative interpretation, since it *does* create
educated black women who can move into the middle class.
quoted me:
No it doesn't "work" - as there now is a gross imbalance in America
between educated middle-class black women and educated middle-class
black men for them to marry,
and replied:
Ed's goal was to have more minority people in the middle class. Of
course, if you are a racist, you believe that sex and marriage between
middle class white men and middle class black women is sinful or some
such fundamentalist nonsense. And I'm sitting here trying to educate
you about statistics.
"Statistics" would tell you that it would still have a huge
imbalance between educated men and educated women - as the total number
OF ALL RACES of men with college degrees is far smaller than of women
of age to marry for the first time.
"Statistics" would also tell you that black-white intermarriage is
nil in America, so your "solution" couldn't work.
No $6 admission! http://stores.ebay.com/INTERNET-GUN-SHOW
.
|
|
|
| User: "Edgar Svendsen" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
25 Sep 2005 04:11:37 PM |
|
|
<editor@netpath.net> wrote in message
news:1127672534.916034.34130@g43g2000cwa.googlegroups.com...
TG wrote:
In terms of Ed's overall goal, indeed, affirmative action works even
under editor@netpath' s negative interpretation, since it *does* create
educated black women who can move into the middle class.
quoted me:
No it doesn't "work" - as there now is a gross imbalance in America
between educated middle-class black women and educated middle-class
black men for them to marry,
and replied:
Ed's goal was to have more minority people in the middle class. Of
course, if you are a racist, you believe that sex and marriage between
middle class white men and middle class black women is sinful or some
such fundamentalist nonsense. And I'm sitting here trying to educate
you about statistics.
"Statistics" would tell you that it would still have a huge
imbalance between educated men and educated women - as the total number
OF ALL RACES of men with college degrees is far smaller than of women
of age to marry for the first time.
"Statistics" would also tell you that black-white intermarriage is
nil in America, so your "solution" couldn't work.
While what you say is true, this thread is kinda about changing what is, for
something preferable. More black women in the middle class will, IMHO, not
be an entirely bad thing, even if there are difficulties for them in
finiding suitable black mates. And, over time, I believe it will tend to
increase the number of interracial marriages which would also be a good
thing. I agree that affirmative action at Universities will not solve all
of the problems of all of the disadvantaged minorities. I don't think that
ought to be the question; I believe that the question is whether the actions
of the Universities is tending to make things better or worse. If the
answer is that they are making things somewhat better,than as Martha Stewart
says "That's a good thing!"
I tend to shy away from solutions that are touted as the best, the optimum
solution. We don't know everything, it's good to have many solutions
tending toward our goals so we can see if there are unforseen side effects
of some of them.
Ed
No $6 admission! http://stores.ebay.com/INTERNET-GUN-SHOW
.
|
|
|
| User: "" |
|
| Title: Re: Affirmative action at 3 univeresities: a report |
26 Sep 2005 02:44:38 AM |
|
|
Ed wrote:
More black women in the middle class will, IMHO, not
be an entirely bad thing, even if there are difficulties for them in
finiding suitable black mates. And, over time, I believe it will tend to
increase the number of interracial marriages which would also be a good
thing.
Black-nonblack intermarriage isn't likely to increase in America.
It also would have many problems if it did - like leaving an
equally-large number of nonblack educated women without possible
husbands to whatever extent nonblack educated men started marrying
black educated women. This is just one of many examples of major
social problems inside the black community in America that only blacks
themselves can solve - like very high illegitimacy rates, rampant VD,
and an enormously-disproportionate murder rate.
Save on gas! Shop the http://stores.ebay.com/INTERNET-GUN-SHOW
.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
