Attachment to letter to John M. Gore

ATTACHMENT TO LETTER OF 11/29/2017 TO ACTING A.G. GORE, CRD, DOJ:

The construction of the SAT distribution curves, broken down to race, are explained in steps [A] to [C]:

[A] Where do the data come from?

(1) The % of applicants to Harvard, broken down to race. Click on * https://college.harvard.edu/admissions/admissions-statistics .

By implication, whites constitute 49.1% of Harvard's applicants that year. Unfortunately, I can't establish the year for which the data applies. But I believe such percentages don't change drastically from year to year.

Ethnicity
African American	14.6%
Asian American	22.2%
Hispanic or Latino	11.6%
Native American or Pacific Islander	2.5%

(2) The mean SAT score of all students in the nation (not just Harvard applicants) in 2015. Click on https://www.insidehighered.com/news/2015/09/03/sat-scores-drop-and-racial-gaps-remain-large

Mean SAT Scores by Race/Ethnicity, 2015

Group	Critical Reading	Mathematics	Writing
American Indian	481	482	460
Asian-American	525	598	531
Black	431	428	418
Mexican-American	448	457	438
Puerto Rican	456	449	442
Other Hispanic	449	457	439
White	529	534	513

Combined SAT Score, and Changes Since 2006, by Race/Ethnicity

Group	Combined Score 2015	Change Since 2006
American Indian	1423	-27
Asian-American	1654	+54
Black	1277	-14
Mexican-American	1343	-28
Puerto Rican	1347	-16
Other Hispanic	1345	-26
White	1576	-6

(3) The national racial characteristics of the math SAT distributions (NOT just for Harvard applicants) . Click on https://www.brookings.edu/research/race-gaps-in-sat-scores-highlight-inequality-and-hinder-upward-mobility/

and https://secure-media.collegeboard.org/digitalServices/pdf/sat/total-group-2015.pdf

You'll see that so far as the Math SAT scores are concerned, the blacks and Hispanics are low-score-heavy when compared with the so called normal distribution (i.e. real Gaussian) while Asians are high-score-heavier and white students show a fairly good normal distribution. I conjecture that the same characteristics apply to "writing" and "critical reading" SAT performances, but much LESS pronounced than the math situation.

[B] The racial characteristic of the total SAT score distributions, described in words :

The following is simulated, based on data described in [A]. But DOJ may soon have the real admissions data to construct the REAL curves. Note: Unfortunately, the maximum SAT score in 2015 is 2400 points, not the 1600 points when Espenshade did his study.

Blacks will have a low-score-heavy Gaussian curve centered on 1277 with a one standard deviation width of about

120, with a height that is roughly 1/3 that of the white (The height difference is owing to the fact that the number of

black applicants to Harvard is only 1/3* that of whites.)

Hispanics will have a low-score-heavy Gaussian curve centered on 1354 with a one standard deviation width of about 130, with a height hat is roughly 1/4 that of the white (The height difference is owing to the fact that the number of Hispanic applicants is about1/4* that of whites.)

Whites will have a fairly regular Gaussian curve centered on 1576 with a one standard deviation width of about 100, with a height that is arbitrarily set at 140.

Asian Americans will have a high-score-heavy distribution curve centered on 1654 with a one standard deviation width of about 170, with a height that is roughly 1/2 that of the whites (The height difference is owing to the fact that the number of Asian Am. applicants is about 1/2* that of whites.)

[C] The racial characteristic of the total SAT score distributions, seen in graphic form :

Again, unfortunately, the highest SAT score in 2015 is 2400 points, not the 1600 points when Espenshade did his study. Hence while Espenshade reported that 140 SAT points were deducted from Asians, on this 2400 maximum SAT scale, it becomes (140 x 2400/1600) = 210 SAT pts.

In the diagram below, the x-axis is total SAT scores of each applicant, and the y-axis is the count of applicants with that particular SAT score. The area below each curve should be equal to number of applicants of each race. I arbitrarily set the max counts to 100. That shouldn't affect the illustrative nature of this diagram, because the height of all other curves are determined relative to whites.

ATTACHMENT Page 6

ATTACHMENT Page 7

The vertical cut-off line of 1850 is again set arbitrarily. Note that the areas to the right of the cut-off line that are under the curves represent the number of students Harvard will admit from each race, if SAT scores were the only standard. Note that at California Institute of Technology, which does not practice "race preference", 45.4% of the admitted students are Asians while the admitted whites are 27.4%.

It is now easy for people to understand why Harvard and other Ivy League colleges will have to take SAT points away from Asian Ams. Without the deduction of points from Asian Americans, Harvard would have to admit MORE Asian Ams than whites, even AFTER legacy, extra-curricular activities, class ranking and letters of recommen-dations from high school counselors are all taken into account. From the above graph, it is also obvious that SAT points must be added to black and Hispanic students in order to fit into Harvard's own definition of what would

Be the proper race balance.

There is a known defect in the above diagram. Only good students apply to Harvard. Hence the peaks of all 4 curves should be shifted more to the right i.e. the high score end, but the tails will NOT. (End)