So now you’re resorting to “appeal to authority” instead of actually answering the question?
I’ll try again:
Are you really going to argue that getting 2-3 extra questions on the SAT makes a student a more worthwhile admit than someone who has extensive credentials outside of class and slightly lesser scores/grades? Say, 2400/4.0 vs. 2290/3.9?
Yes. The one with 2400/4.0 is OBJECTIVELY academically better. Notice the capitalized word? And you’re going to get the best academic results possible by accepting the best academically prepared students possible.
How certain are you that you know what Oxford’s web page means by “academic abilities and potential”?
This post confirms that you’re actually a little shaky concerning what the word “academic” means to academics!
Except it doesn’t. Oxford expects certain minimum scoring levels but does not guarantee admission based solely on high scores. If we had the equivalent of A or O-levels in the US, you might have a point, but we don’t.
The entrance exams you take for Oxford (or schools in other countries) tend to cover a wide range of subjects you take in high school, which is significantly different than the SATs, which are not designed to actually test mastery of any particular classes but rather more general knowledge.
I’m not even sure how Terr thinks it makes sense for Oxford to say they’re using an interview to determine an objective measure of some kind of test score (or something). Why an interview, Terr?
I am sure you think that “academic” means “heavily involved in his community”, “has glowing recs”, “has devoted much time and energy to hobbies of passion”, and “has an actual personality”.
Oxford’s criteria for admissions are academic excellence. Not being “interesting”, or “overcoming adversity”, or being “passionate”, or “having tough background” or having the right skin color.
Note: I never said -300 was 2-3 questions. I simply claim that the number of questions across your idealized thresholds is too small to count for anything significant.
For the sake of fairness, let’s get specific, then.
http://www.sat-tutors-blog.com/2009/01/23/489/
Let’s assume a fairly uniform drop in each category.
Dropping from 2400 to ~2300 = about 770 in each category
Raw score is calculated: Correct answers -.25Wrong answers
C-.25(T-C) = Raw, where C=Correct, T=Total, and W=Wrong=(T-C)
Raw score CR = 63/67
Raw score in Math: 52/54, roughly
Raw score in Writing: We’ll go with 48/49, roughly
770+760+780 = 2310, which gets lumped in with 2300’s in the end anyway.
So:
C-.25*(67-C) = 63, C=63.8, W=3.2
C-.25*(54-C) = 52, C=52.4, W=1.6
C-.25*(49-C) = 48, C=48.2, W=0.8
Sum(W) = 5.6 questions wrong, across three tests, or about 1.9 questions per test, drops you from 2400 to 2300.
How about to 2100, a grand change of -300?
Choosing raw scores of 58, 48, and 44 (exactly a 2100 score), the same calculations give us
C-.25*(67-C) = 58, C=59.8, W=7.2
C-.25*(54-C) = 48, C=49.2, W=4.8
C-.25*(49-C) = 44, C=45.0, W=4.0
Sum(W) = 16 questions wrong, across three tests, or about 5.3 questions wrong per test, drops you from 2400 to 2100.
Limited, but not so limited that the usefulness of the metric is conserved. The coachability of the tests weaken their strength considerably and it’s completely silly to pretend like it doesn’t. It screams of ignorance of the history of the test and the history of past performance.
As I just showed you you can improve from a 2100 to a 2300 by just getting 3-4 extra questions right per section. Does that sort of jump really scream worlds of difference to you?
Except a 2400/4.0 is not necessarily objectively better. A test taker can get a 2400 in one session and a 2280 the next by sheer variance alone. Having a high SAT score does not guarantee that your student is the best academically-prepared student possible.
I should also note that top colleges superscore the SAT, so you could theoretically get 800/200/200, 200/800/200, and 200/200/800 in three separate sessions and get a 2400.
And yes, like others noted, Oxford’s admissions is not 100% objective, either. It’s not as focused on diversity as the top unis here, but it’s by no means fully “objective.”
Because an effective policy would have to have a number of conspirators acting in coordination. Things on that scale are usually hard to hide.
Cite? I don’t beleive there has been a dramatic jump in Asians at either Harvard or Princeton, the two schools under investigation.
It was the continuation of a trend that had been going on coupled with more spots being “taken away” from under-represented minorities. Whites were basically unaffected. Berkeley first used race-blind metrics in 1998. From 1996-1998, Asian enrollment was 36.4%, 38.5%, and 39.5%. The changeover could only be responsible for a 1% increase. That increase was smaller than the change between the two years immediately preceding it. For Whites, the numbers were 29.1%, 28.3%, and 28.2% respectively. What did change was the underrepresented minority percentages, which were 22.2%, 22.4%, and 11.7%. So the shift was mostly from Hispanics (and some Blacks) to Asians.
This is largely not true. If you have some case you would like to present, I would love to see it.
You say that now, yet you have accused a number of people in this thread of racism and characterized the process as racist. If you think such accusations do not connote malice, then we will have to agree to disagree. It’s not a strawman, you actually said these things. And although you may be changing your tune now, the miscommunication is mostly because you are constantly moving the goalposts.
Let’s assume for a second that is true. Why is that any worse than pitting Maryland applicants versus one another? Maryland consistently has much better schools than most of the US, which means intra-Maryland competition makes it harder for individuals. Is that wrong? What about international students competing against one another? Or basketball prospects competing against one another?
You’re not? You asked me:
My response was to that comment. The point is that his claim doesn’t pass a basic smell test.
You can’t unless you have ALL the metrics used. Let’s use a basic analogy. Every year, the NFL has a combine to showcase and measure all the prospective draftees. So they figure out how strong you are, how fast, and how high you can jump. They do interviews, they give an aptitude test, etc. Now, let’s say I gather all those numbers, and find out people who have a high scores on the aptitude test are three times less likely to be drafted. Would it be fair to assume that NFL scouts are biased against smart people?
What you seem to be assuming that such a negative correlation, between being Asian, and getting into an elite university must mean substantial discrimination. The problem with that logic is that you have no idea how such admissions decisions are made. The relatively likelihood you mention is based on the factors you are aware of, not all of those that are used. Since you don’t know the impact they have on the overall decision making process, you have no idea how unfair the process is. This is further complicated by a self-selection bias given that only generally qualified people bother to apply, meaning your applicant pool is far more uniform, far more clustered quantitatively, and far less differentiable. So there are less bases for comparison, meaning that even if, for example, Asian have higher SAT scores and more AP classes, the relative value of those things could be minimal in differentiating between the capable and the incapable. As such, a college may only use those factors (eg. SAT score, etc), which you use to substantiate discrimination, and baseline above which a certain score has no more value. If that is the case, and there evidence there is, then your relatively likelihood is not accurate.
So you equivocate a test result with the adjective “academic”. Afterwards, you make the questionable conclusion that if we select the best academically prepared students (in your definition) we will get the academically best results (in your definition). Even if we accepted your equivocation—and we won’t—the conclusion still doesn’t follow. Even if test scores are the best predictor of whatever, this is obviously based on which people are selected in the first place by the very boards being called into question. In other words, it is not a reasonable inference from statistics to conclude that we will get the best “academic” anything if we change the way the selection process works.
You guessed wrong. To an academic, “academic ability” and “academic potential” involve (at the least) intellectual curiosity, creativity, a desire for depth of understanding, and the valuing of rigorously developed positions.* I’d like to see the test that measures these characteristics, such that a computer could score it and all informed people would agree the computer was correct in each instance. (There isn’t one.) These are the kinds of things Oxford is hoping to see in an interview. What did you think they were hoping to see in an interview?
Note that the characteristics you named all could easily be indicators of these, as can test scores. But none individually is a sure indicator of them. It’s a holistic judgment.
Hasn’t it been long established that GPAs and SAT scores have no predictive power when it comes to success at the college level? Maybe I’m remembering wrong.
If by “not as focused on diversity” you mean “not at all focused on diversity” you’re right.
You’re remembering wrong. Upthread I posted an article that shows that SAT scores are a very good predictor of success at college level.
In addition, 4.0 isn’t even OBJECTIVELY better. 4.0 is either
A) A weighted GPA. My weighted GPA was around 4.8, so 4.0 is a pretty bad weighted GPA.
B) An unweighted GPA, in which case it doesn’t take into account how many honors courses you’ve taken. A 4.0 GPA with all honors courses is OBJECTIVELY better than a 4.0 GPA with no honors courses.
Terr, the problem is that you’re putting way too much stock into metrics that are only part of what we’d wish to analyze in assessing the quality of a student for admission.
The whole point of holistic admissions is to take multiple components into consideration.
Colleges want to admit strong-minded people with lots to contribute. They aren’t too interested in admitting study-bots.
“Holistic” admissions are subjective and opaque, ripe for abuse that results in discrimination. And besides that, if they are truly “looking at you as an individual”, that has nothing to do with “diversity” that the universities reportedly seek. Because assessing people as individuals does not achieve diversity. Diversity is achieved by looking at people as members of groups.
You’re making a wild assumption that has no real evidence to support it. The evidence you to point to is weak and not very compelling when you consider the rest of the relevant context.
Oh, yeah:
Need I also remind you that the SAT has a subjective component to it – the written essay of the writing section? The essay is graded on a 12-max scale across two readers. Still objective?