<p>Well bluebayou, it does make life simpler.</p>
<p>One thing you have to consider is that universities (and people at them) can prepare multiple narratives for telling the story of the place. Facts that are equally true can tell conflicting stories–which are also equally true! </p>
<p>Depending on the purpose and the audience, you can talk about how research demands or budget cuts are decreasing undergraduate access to faculty. Or you can talk about how you’re maintaining or improving those same factors. Where one department is being cut, another is adding staff. Where one research grant is used to buy out a teaching obligation, another is used for increasing undergraduate research participation. An institution will tell legislators how appropriation cuts could devastate academic quality, while at the exact moment people on campus are working on plans to make sure those cuts will have a minimum impact.</p>
<p>Grabbing quotes or selective data help tell a great story about an institution, but they don’t tell the whole story. They’re valuable but you have to also consider the other perspectives. I think most of us grasp that when it comes to USN&WR-compiled statistics (they don’t tell the whole story of the place); we should also bring a little bit of interpretive caution to blogs, press releases, and news articles about institutions.</p>
<p>In case you are interested, the weights for the 12-factor model that includes National Academy of Sciences members and which corrects for the Chicago/Berkeley prediction errors are:</p>
<p>intercept 137.57852
SAT CR 75th .04011
yield -.00446
research $ percent .00418
public or private -.17452
SAT math 75th squared .0000029
graduation percent squared .00010944
number of freshmen squared .000000003138408
SAT CR 75th natural log -25.37327
endowment per FTE natural log .04336
number of freshmen natural log .21825
number of NAS members .01113
NAS members squared -.00005145</p>
<p>For Liberal Arts Colleges, the Peer Assessment rating is also very predictable using IPEDS data. I obtained a model of PA that explained 91% of PA ratings. The multiple correlation was .95. This is extremely high.</p>
<p>I used the following factors or various mathematical transformations of them:
women or coed
number of freshmen
number of bachelors degrees
instructional $ percent
public or private
SAT CR
SAT math
admit percent
endowment per FTE</p>
<p>rank based on actual PA, actual PA, predicted PA, difference, school</p>
<p>1 4.7 4.5 0.2 Amherst College
2 4.6 4.5 0.1 Swarthmore College
3 4.6 4.7 -0.1 Williams College
4 4.4 4.5 -0.1 Wellesley College
5 4.3 4.2 0.1 Bowdoin College
6 4.3 4.1 0.2 Carleton College
7 4.3 4.0 0.3 Grinnell College
8 4.2 3.8 0.4 Davidson College
9 4.2 4.2 0.0 Middlebury College
10 4.2 4.7 -0.5 Pomona College
11 4.2 4.1 0.1 Smith College
12 4.2 4.3 -0.1 Vassar College
13 4.2 4.1 0.1 Wesleyan University
14 4.1 4.0 0.1 Harvey Mudd College
15 4.1 4.0 0.1 Haverford College
16 4.1 4.1 0.0 Oberlin College
17 4.0 3.7 0.3 Bryn Mawr College
18 4.0 4.0 0.0 Claremont McKenna College
19 4.0 3.7 0.3 Macalester College
20 4.0 3.6 0.4 United States Naval Academy
21 3.9 3.9 0.0 Barnard College
22 3.9 3.8 0.1 Colby College
23 3.9 4.0 -0.1 Colgate University
24 3.9 . . United States Military Academy
25 3.8 3.8 0.0 Reed College
26 3.8 4.2 -0.4 Washington and Lee University
27 3.7 3.7 0.0 Bucknell University
28 3.7 3.7 0.0 Colorado College
29 3.7 3.5 0.2 Kenyon College
30 3.7 3.6 0.1 Trinity College
31 3.6 4.0 -0.4 Hamilton College
32 3.6 3.5 0.1 Occidental College
33 3.6 3.7 -0.1 Scripps College
34 3.6 3.3 0.3 Sewanee: The University of the South
35 3.6 3.5 0.1 St. Olaf College
36 3.6 3.8 -0.2 University of Richmond
37 3.5 3.2 0.3 Centre College
38 3.5 3.5 0.0 College of the Holy Cross
39 3.5 3.5 0.0 Connecticut College
40 3.5 3.5 0.0 Furman University
41 3.5 3.3 0.2 Rhodes College
42 3.4 3.3 0.1 Berea College
43 3.4 3.3 0.1 DePauw University
44 3.4 3.3 0.1 Dickinson College
45 3.4 3.2 0.2 Earlham College
46 3.4 3.5 -0.1 Franklin and Marshall College
47 3.4 3.7 -0.3 Lafayette College
48 3.4 3.5 -0.1 Skidmore College
49 3.4 3.0 0.4 Spelman College
50 3.3 3.3 0.0 Agnes Scott College
51 3.3 3.5 -0.2 Denison University
52 3.3 3.4 -0.1 Gettysburg College
53 3.3 3.0 0.3 Wabash College
54 3.3 3.4 -0.1 Wheaton College
55 3.3 . . Whitman College
56 3.2 3.1 0.1 Beloit College
57 3.2 3.1 0.1 Hendrix College
58 3.2 3.4 -0.2 Kalamazoo College
59 3.2 3.2 0.0 Lewis & Clark College
60 3.1 2.8 0.3 Birmingham Southern College
61 3.1 2.9 0.2 Cornell College
62 3.1 2.9 0.2 Drew University
63 3.1 3.0 0.1 Goucher College
64 3.1 3.1 0.0 Gustavus Adolphus College
65 3.1 2.9 0.2 Hope College
66 3.1 2.8 0.3 Luther College
67 3.1 2.9 0.2 Saint Johns University
68 3.1 3.2 -0.1 Willamette University
69 3.0 3.1 -0.1 Austin College
70 3.0 3.3 -0.3 Illinois Wesleyan University
71 3.0 2.8 0.2 Millsaps College
72 3.0 2.9 0.1 Ohio Wesleyan University
73 3.0 3.2 -0.2 Southwestern University
74 3.0 2.8 0.2 University of North Carolina at Asheville
75 3.0 3.3 -0.3 University of Puget Sound</p>
<p>Objective Criteria Utilized:
</p>
<p>Results:
</p>
<p>xiggi’s past statement:
</p>
<p>Now, xiggi, what “discrepancy” are you talking about? 
</p>
<p>^^perhaps xiggi really wants to compare the PA’s of CMC and Smith? :rolleyes:</p>
<p>Hmmm. No wonder Xiggi doesn’t like PA. CMC’s score is an embarrassment.</p>
<p>Hmmm…[Departments</a>, Programs, Majors - CMC](<a href=“http://cmc.edu/academic/departments.php]Departments”>http://cmc.edu/academic/departments.php)</p>
<p>Departments</p>
<p>Government
History
Joint Science
Literature
Mathematics
Military Science
Modern Languages
Philosophy
Religious Studies
Physical Education and Intercollegiate Athletics
Psychology
The Robert Day School of Economics and Finance </p>
<p>That’s it?</p>
<p>"Majors & Programs</p>
<p>4 + 1 BA/MBA Program
Accounting
American Studies
Asian-American Studies
Asian Studies
Biology
Biology-Chemistry
Black Studies
Chemistry
Chicano Studies
Computer Science
Economics
Econ-Accounting
Environment, Economics, and Politics
Environmental Science
Ethics
Financial Economics
French
Freshman Humanities Seminar
Film Studies
Gender Studies
German
Government
History</p>
<p>International Relations
Joint Music
Korean
Leadership
Legal Studies
Literature
Management Engineering
Mathematics
Military Science
Neuroscience
Philosophy
Philosophy and Public Affairs
Philosophy, Politics, and Economics
Physics
Practicum
Psychology
Religious Studies
Robert Day Scholars Program
Robert Day School Master’s Program in Finance
Science and Management
Senior Thesis
Spanish
Speech
Theatre"</p>
<p>That’s it?</p>
<p>Berkeley offers some 300 degree programs, listed here in alphabetical order. Also see the Division of Undergraduate Education, the Graduate Division and Undergraduate Affairs. </p>
<p>[Academic</a> departments & programs - UC Berkeley](<a href=“http://berkeley.edu/academics/dept/a.shtml]Academic”>http://berkeley.edu/academics/dept/a.shtml)</p>
<p><a href=“http://are.berkeley.edu/teaching/undergradprogram.html[/url]”>http://are.berkeley.edu/teaching/undergradprogram.html</a></p>
<p>not fair, dstark. A CMC student can always stroll over to the higher-ranked Harvey Mudd and take classes there. :)</p>
<p>lol… that’s right…but a Claremont kid can’t compete with a Harvey Mudd kid. The Harvey Mudd kid probably has higher SAT scores. :)</p>
<p>CMC’s PA is way, way too high.</p>
<p>When I look at it’s massive amount of departments…you have to take out military history…phys ed…well all of them except government, econ and finance (which is one department), literature and history.</p>
<p>The school barely has any departments of its own.</p>
<p>And you have to take away the government department. Government can’t do anything and the school has a whole department on it.</p>
<p>Some of you guys are hilarious. Absolutely hilarious!</p>
<p>:)…</p>
<p>
</p>
<p>Au contraire. The (California) Government can raise taxes to help Cal keeps it’s inflated PA. </p>
<p>Perhaps the CMC kid should stroll over to Scripps – kinda like Smith-West?</p>
<p>glad we could provide you with today’s smile, xiggi!</p>
<p>Oh I found the Dstark/Bluebayou routine irrestible and humorous. Perhaps a bit less humorous than reading that women versus coed, the number of freshmen, the number of bachelors degrees, public vs private were core elements of the PA correlation, and seemingly dwarfed SAT CR, SAT math, admit percent, and endowment per FTE sufficiently to arrive at the projected results.</p>
<p>xiggi-
women vs coed had a miniscule relationship with LAC’s PA but it apparently captured something the other factors did not, which was probably the tendency by the experts to give women’s colleges a liitle boost in PA because they are women’s colleges. </p>
<p>The individual correlations with PA were as follows:
.75 SAT math 25th
.72 SAT CR 25th
.71 percent of expenditures spent on instruction
.71 SAT math 75th
.66 SAT verbal 75th
.61 graduation percent
.56 endowment per FTE
.55 retention after 1 year
.45 academic support expenditures percent
.11 number of bachelors degrees
.04 number in freshman class
.03 yield
.02 public or private
.001 public service expenditures percent
.0009 women or coed</p>
<p>Previously, I had not listed them in order of importance.</p>
<p>^
</p>
<p>
</p>
<p>Now THAT is humorous.</p>
<p>When I posted my comment about the LAC model, the only information shared was a listing. No correlations - no weights!</p>
<p>
</p>
<p>My dear friend UCB, perhaps you’ll find equally humorous to seek the answer to one of your own questions. Earlier, you asked me “Now, xiggi, what “discrepancy” are you talking about?” to (I shall assume) deride my previous comments about Harvey Mudd and Smith College. </p>
<p>As you probably know, I have been drawing parallels between the non-coed schools and Harvey Mudd or Pomona for a number of years. Perhaps, you could go back to the 2007 and 2008 versions of the USNews as well as past discussions on CC to understand the comparisons between the schools’ PA and the listed selectivity elements. </p>
<p>In this case, would you look at the SAT scores, admission statistics, and PA for the years past and tell me how the examples of Smith College and Harvey Mudd look … and espouse the models of CollegeHelp? Any comments regarding the SAT scores? Any comment why Smith has a PA of 4.3 and Harvey Mudd 4.1? Is it because of “.75 SAT math 25th, .72 SAT CR 25th, .71 percent of expenditures spent on instruction, .71 SAT math 75th, and .66 SAT verbal 75th?”</p>
<p>For instance, in USNews 2007, you would have found the following mumbers</p>
<p>Harvey Mudd College (CA)<br>
PA 4.1<br>
Selectivity Rank 1<br>
SAT 1420-1550<br>
Top 10% 94%
Admission Rate 30%</p>
<p>Smith College (MA)<br>
PA 4.3<br>
Selectivity Rank 46<br>
SAT 1140-1370<br>
Top 10% 61%
Admission Rate 53%</p>
<p>By the way, Pomona (the 3d most selective LAC in the country) had a PA of 4.2, but Grinnell earned a 4.3. Plug the SAT numbers in!</p>
<p>By the way, the analysis of LACs was based on about 150 LACs listed in US News National Liberal Arts Colleges with SAT CR and SAT math both over 450.</p>