<p>Why do SATs generate such strong emotions? That is an interesting question by itself.</p>
<p>I don't think there is anything wrong with my calculations.</p>
<p>The confusion comes from focusing on individual students instead of the larger university picture. There are many factors that affect whether an individual student succeeds: homesickness, drug and alcohol abuse, mental health issues, a change in interests, family problems, financial problems, difficult major, demanding faculty. The SATs can't account for the things that happen with students after they get to college. Consequently, students with the same SAT score don't do equally well. There is individual variability among students with the same SAT score.</p>
<p>To control for this noise, you can calculate the average gpa or the graduation percent for groups of students with the same SAT score. Then you can correlate SAT scores with the average gpa and graduation percent for groups with the same SAT score. When you group the data this way, you are left with the "pure" relationship between SAT and grad rate or gpa.</p>
<p>If a university wants to improve its graduation rate, it should enroll students with higher SAT scores. On average, students with higher SAT scores are more likely to graduate. You can simply inspect the data in US News Best Colleges and see the relationship with your naked eyes.</p>
<p>What I did in my earlier post was very simple. Anybody willing to learn how to navigate the IPEDS website could replicate what I did. Simply download data for a range of colleges for SAT and graduation rates and calculate the correlation. Use any software you'd like.</p>
<p>Collegehelp, I don't know if your conclusion is correct. It may be on an overall basis, looking at all schools, but in a given school, especially the selective ones, that has not been the case. Grades, which reflect work ethic, seem to be the biggest indicator. Again, I have not done a statistical analysis of this.</p>
<p>If a university wants to improve its graduation rate, it should enroll students with higher SAT scores. On average, students with higher SAT scores are more likely to graduate. You can simply inspect the data in US News Best Colleges and see the relationship with your naked eyes. collegehelp.</p>
<p>I strongly suggest you grab a hold of the study done by Wake Forest University PRIOR to their decision to go SAT optional (not dropping it...just making it optional for those students who OTHERWISE qualify for admission and yet have SAT scores on the bubble.) </p>
<p>They have scientific proof, they say, that there is NO correlation between SAT scores and graduation rates or success in college. Furthermore, they have admitted kids in the past with a very good application file but with more moderate scores who have excelled and gone on to GREAT things in Graduate School and Professional Schools. </p>
<p>No, ma'am, I am not suggesting no standardized testing. I am a vehement opponent to this insidious ranking and "prestige" shopping that goes on and people who represent one college is "so much better than another college" because SAT scores are higher. That is utterly false. And a disservice to ALL students, those blessed with a HIGH SAT and those with more modest scores.</p>
<p>Most admissions officers tell us they seek to formulate a "class" and not just a gathering of kids with perfect SAT scores. Standardized testing is but one factor and a narrow one at that. </p>
<p>Those colleges that are now SAT optional, are saying they will consider them if submitted but not punish kids who dont submit them. They will look very deeply at their high school transcript and get a better picture of the competitiveness of that school, the courses offered and the level of teaching quality there. They will look at creative factors and the whole person. It makes their job "harder" to quote Wake Forest, but more just and perhaps get a better class. They anticipate actually having a very high quality of admitted students. Just not measured by a test on one saturday morning.</p>
<p>They would rather have a kid with exceptional qualities, high grades, high ambition and motivation and WELL ROUNDED than a kid with simply the highest SAT score.</p>
<p>If someone wants to go to Grinnell or Kenyon or Emory or NCState they should select that school based on fit and their personal interests, not on its so called "prestige" ranking in USNWR. There are plenty of kids with high scores and high gpa's at most first and second "tier" schools.</p>
<p>That is all I was saying. I recognize that a lot of people on CC are strong proponents of prestige rankings and using SAT scores to separate students and pit one college against another. There are plenty of threads on that basis. I simply disagree and oppose that notion.</p>
<p>nocousin-
I am sure that Wake Forest can point out some students with mediocre SAT scores who became successful. I am sure there is an explanation for these idiosyncratic outcomes. But, their overall conclusion is incorrect. Wake probably wanted to justify going SAT-optional and came up with an excuse (ala "weapons of mass destruction").</p>
<p>cptofthehouse-
I know of one school that did the very thing I suggested in post #41. In fact, my ideas about this came from that internal study. They classified engineering students according to math SAT level. They calculated the graduation rate FROM ENGINEERING (not transferring to an easier major) for groups of students who had the same SAT score. The correlation between SAT and graduation percent was +.87 which was very high. This is within a college, not among different colleges.</p>
<p>
[quote]
Those colleges that are now SAT optional, are saying they will consider them if submitted but not punish kids who dont submit them. They will look very deeply at their high school transcript and get a better picture of the competitiveness of that school, the courses offered and the level of teaching quality there. They will look at creative factors and the whole person.
[/quote]
</p>
<p>I generally find that those that did well on the SAT think it is a worthwhile measurement and those that didn't do well don't think it is a worthwhile measurement.</p>
<p>I have a further problem with the Wake Forest approach in that it forces students to try to game the system and decide if having an SAT score 40 points below the median is too low to submit or not. That kind of decision should not have to be part of the candidate's application. </p>
<p>Why would Wake want less info on a candidate rather than more? If they accept the SAT, then have everyone submit it and the school can decide how relevant it should be for each individual candidate. However, putting students through this game theory is ridiculous and basically the school is being hypocritical trying to pretend they don't have lower SAT scoring students at their school that otherwise would be accepted so that its reported SAT averages don't go down when they accept them. They are basically encouraging candidates to be devious but not too devious with their applications. This is the main form of creativity that I see manifested in their applications.</p>
<p>They want more app is one reason. The other reason is that there are kids who are excellent students who don't test well. They may not bother to apply with the odds against them. It is not completely a "game". They do have some altruistic reasons for the policy change. However, I don't expect to see Wake's SAT scores dip from this change in policy.</p>
<p>Well, maybe I'm an outlier, but I did very well on the SAT back in my day and my D has done very well on hers, to the point these scores will help her immensely in college admissions. But I remain deeply skeptical of their value. </p>
<p>Everything I've read suggests:
1) Girls on average score about 30 points lower than boys in the SAT, yet girls on average consistently outperform boys in both HS and college grades. This includes math, where boys' SAT scores are on average significantly higher than those of girls who earn the same grades in the same college math classes.
2) SATs consistently under-predict the college performance of bilingual students, including bilingual students whose best language is English. A University of Maryland study concluded that Hispanic students there on average scored about 90 points lower on the SAT than their non-hispanic white counterparts, but on average earned the same grades.
3) SAT scores generally do a very poor job of predicting the college performance of older (non-traditional) students.
4) Even the College Board admits that HS grades are a better predictor of college grades than the SAT.
5) A University of California study found that SAT II scores were a better predictor of first-year grades than the SAT I, and in addition SAT II scores are less sensitive to the test-taker's socioeconomic background than SAT I. At that point the UC President proposed eliminating the SAT I in UC admissions, reasoning that it "added very little" to an assessment based on HS grades and SAT II scores. That proposal was never implemented, but it did force the College Board to revamp the SAT I. The new SAT I, however, proved to be not much better than the old one, although some studies have concluded that the SAT I writing component is a slightly better predictor of college grades than the math or critical reading sections.
6) Some studies have concluded that SAT scores have a stronger positive correlation with family income and parents' occupations (managerial/professional = high SAT scores) than with college grades. The UC study concluded that when you control for these socioeconomic factors, SAT I scores alone have very little predictive value as to college grades.</p>
<p>bclintonk-
Females generally score lower on the SAT because of the math section. They are not as good at math, on average, but they are better than males in verbal skills. They graduate at higher rates than males because males develop more bad habits once they get to college. Maybe it is the testosterone. If you did the analysis separately for males and females, I think you would control for one of the extraneous factors (work ethic?) and you would see a relationship for females between SATs and graduation rate. Likewise for males.</p>
<p>Regarding students whose English is their second language. It is not surprising that they don't do as well on the CR part of the SATs. I believe Hispanics have lower graduation rates. The Maryland study you quoted...I'd like to read how they did it. I am skeptical.</p>
<p>I think HS grades are useful because you can see which grades the student received in which subjects and the level of difficulty of the coursework. HS gpa and HS rank are highly correlated with SAT. They might be better predictors statistically but not by much. But, SATs are standardized. Grades are not. Some high schools are more competitive than others. HS rank and gpa are relative.</p>
<p>When SATs are correlated with family income it is because (a) smarter parents make more money and (b) smarter parents have smarter kids. The UC study is a good example of poor social science research. If you control for income, you control for parental ability. It is like saying that every student has about the same ability after you subtract the ability of their parents. The UC study misunderstood the causal pathways among the factors they studied.</p>
<p>SAT I scores are very good at predicting college grades. I'll try to find evidence of this. I've seen it before.</p>
<p>Lots of generalizations floating around here that I dont agree with. SAT skills can be coached and very often are at private prep schools. Some kids dont score as well, but when you examine their test results its because they moved slower not because they got questions wrong. The "gaming" that goes on is with the people who get lots of help preparing for the exams.</p>
<p>They are one factor of many in admissions decisions. Its not hard for a student applying to Bates, Holy Cross or Wake to determine what the 50th percentile is for admitted students and decide if their scores are on the bubble or not and whether or not to submit them. They may indeed get more apps, but maybe those new apps will be from highly intelligent and creative kids who otherwise would have been discouraged from applying and who would bring a lot to the table and become very productive students at those schools. </p>
<p>If you did well on the SAT, great. Submit your scores. In selecting a college a student would be wise to examine a number of factors including what type of student goes there and whether they fit into that "club" or not. Fit is the most important factor, it seems to me on success in college. But this business of "my school is better than yours because our SAT scores are higher" is really silly and frankly not the kind of school where I want to send my kid anyway if that is a pervasive culture. Every school in the United States serves a purpose and we should applaud the students at ALL of these schools, whether top tier or fourth tier. </p>
<p>My D is near the very top of her class at her college and she is plenty challenged and thriving. There are plenty of kids there with uber high stats and plenty of kids with more modest SAT scores comparatively. Her overall experience has been wonderful. </p>
<p>Except for the kid majoring in Astro Physics at CalTech or MIT or whatever, most kids would be just fine at any school, and their success in life is not going to be hinging on whether they got into Princeton or not or whether the avg SAT score at Cornell is higher or lower than the SAT scores at BC. That is insidious. And odious. JMHO.</p>
<p>Nothing wrong with getting help with the SATs. I recommend it, as a matter of fact. Problem comes from fixating on the results. Sons applied to schools all over the map in terms of SAT range and his favorites were not the ones with the highest ranges.</p>
<p>Except that the schools that have looked at this generally find that females do just as well as males in college math courses, despite having lower SAT scores. In other words, males are better at SAT math but not better at college math. The SAT incorrectly predicts that males will do better in college math courses, and in so doing it systematically disadvantages female applicants.</p>
<p>As for Hispanic graduation rates, it's true that Hispanics graduate college at lower rates than whites but that's largely a function of income. Low-income whites, blacks, and Hispanics graduate at lower rates than higher-income members of those same groups. You seem to want to claim this reflects a genetic intelligence factor. Apart from the fact that this is a racially loaded claim given the income disparities in our society by race/ethnicity, I think there's plenty of reason to be skeptical. A far more plausible explanation, in my view, is that on average financial aid tends to be inadequate for low-income students, who must work more hours and in many cases take "breaks" in their education to support themselves. Some significant fraction of those who leave never get it together financially or otherwise to come back.</p>
<p>As for the UC study, I don't recall their having said anything about causation one way or the other. They merely said the apparent predictive success rate of the SAT could be accounted for almost entirely by income; generally, the higher the income, the higher the SAT scores and the higher the college grades. But if that's the case, then the SAT is pretty redundant with FAFSA. If all they cared about was who would get the highest grades, they'd have about as much success relying on FAFSA family income data as they have relying on the SAT, which is to say, a little, but not that much---not nearly as much, they concluded, as a combination of high school grades and SAT II scores. </p>
<p>And of course they knew, as everyone in the business knows, that HS grades are not standardized. But despite the fact that they're not standardized, they're STILL a better predictor of college grades than the SAT. Even the College Board admits that.</p>
<p>Cpt: True. I didnt criticize getting help per se. I criticized that the vast majority who get help are the privileged few at private prep schools, some of whom have CLASSES on the SAT (and some of them that keep it off the transcript to hide it). It becomes a class/elitist issue then. But I also suggested that since the SAT can be "gamed" by prep classes, that it skewers the results as well. And therefore the value of the SAT exam.</p>
<p>That was ALL cited by Wake Forest in their studies. Bates, Holy Cross and Wake Forest should ALL be applauded for trying to make the admissions process more just and fair. Will their making the SAT optional bring about the desired results? I dont know. They are learning as they go as well. But many schools were SAT optional many,many years ago and are in fact returning to those days. Wake is one of those schools.</p>
<p>Based on information from a group of 3 schools, here is how different majors ranked in SATs. I don't think this pattern would necessarily be the same at all schools.</p>
<p>Math SAT
computer science
math
physics
computer engineering
economics
software development/programming
electrical engineering
mechanical engineering
chemical engineering
chemistry
industrial engineering/operations research
computer technology
biology
poly sci/international relations
business
art/graphic design
communications
psychology
criminal justice</p>
<p>bclintonk-
What I learned in college was that verbal/math intelligence was more than 50% hereditary based on identical twin studies. Perhaps you were text-messaging during that lecture? :)</p>
<p>Interesting that a math major would be so high for math SAT but lower than physics or even economics for verbal SAT. Where do majors like English, history, philosophy fit in. What group of schools are these lists from?</p>
<p>^^ I don't dispute that verbal/math intelligence is largely hereditary. I do dispute the claim that the SAT is an accurate, culturally neutral measure of it.</p>