College Panel Calls for Less Focus on SATs
nytimes.com
nytimes.com
Truth be told, the SATs aren't particularly hard so most bright people could walk in and take the test cold turkey and pull at least a 1500. Of course, there were a lot of false positives from people who took prep courses and worked their asses off (but, as a former princeton review tutor, I never saw more than a 100 point gain in any student no matter how hard they worked - and most were lucky to get 50 points). A kid whose parents had a few grand to burn on a princeton review course could boost his score from 1300 to 1350, but never to 1550.
The problem with GPA/extracurriculars were that a lot of not very bright people could pull 4.0s and have good 'leadership' experience by gaming the system. And, even worse, a lot of smart people couldn't. A lot of the smartest people I know barely graduated highschool (or didn't). They saw the bullshit for what it was, and often had a knack for rubbing people the wrong way (especially at 16 - most of them can 'play the game' as adults).
(Note: all my experience was with the old 1600 point SAT)
Published research agrees with this:
http://psychology.uwo.ca/faculty/rushtonpdfs/2006%20Intell%2...
Although the SAT is widely believed to measure mainly academic achievement, Frey and Detterman (2004) have shown that it is an excellent measure of general cognitive ability (g). For example, they found that scores on the SAT correlated highly (0.82) with g extracted from the Armed Services Vocational Aptitude Battery (ASVAB), as it also did with g loaded mental ability tests such as the California Test of Mental Maturity (0.82), the Otis–Lennon Mental Ability Test (0.78), the Lorge–Thorndike Intelligence Test (0.79), the Differential Aptitude Test (0.78), the Henmon– Nelson Test of Mental Maturity (0.65), the Coop School and College Ability Test (0.53), and the Raven's Ad- vanced Progressive Matrices (0.48; 0.72 when corrected for restricted range).
Many college admins view this as a feature, not a bug. I've met a few, they tend to be that type of person themselves. It's a natural human trait to try to raise the status of people like yourself.
Incidentally, I went from about 1250 to 1450 with a couple months of prep (no classes, just some books). I'd be curious if you know of good statistics out there.
Is that a "false positive" from a school's point of view? The goal is to predict who will do well in college. You can do well in college by being smart (quick comprehension, good memory, good brain for math, etc.) or by studying hard.
Are there MIT graduates who are not smart?
I'm unhappy about the weight SAT (and ACT) scores are given in the application process because I will be a (junior level) transfer student and I see even less relevance of the SAT in my position when it's already fairly useless for freshmen. I find it mindboggling and frustrating to think that half the colleges I am applying to are requiring such a useless thing while the others are doing the right thing by not even caring for such a score.
In college, I have taken math classes above second semester calculus. I have fulfilled basic general education requirements. I have fulfilled more than expected of a junior-level transfer student in classes related to my major. All of those credits should overwhelmingly trump what I can do in a few hours with a pencil and a calculator for something laughable in comparison. To the point that taking such a test would be unnecessary to point out what I am capable of doing: honestly, is how well I do on some trivial algebra (lazily done entirely with my ti-89) or writing going to tell you how appealing of a computer science major I will be? I'm pretty sure you can safely ignore the SAT and take a look at my classes, grades, experience and recommendations to make a better guess. And the same should apply to high school students as well.
Not to say I did not do well on the last SAT exam I took (on the contrary, because the material is so basic), but why I had to spend the money and waste the time to do so in the first place just stinks. I have taken prep classes way back in high school, and I've tutored people in turn. The article is spot on when it says that students are spending time gaming the SAT.
(On a slightly different sidenote, I don't think the SAT is the only cause for high school classes (at least in the US) being so focused on a small curricula, I believe that's more directly caused by NCLB and tying school performance on exams to funding. But that's a whole different rant/issue in itself.)
Generally, the SAT is administered to high-school students, not college students. It's purpose is to predict the performance of those high-school students, if and when they are in college. The point of any test is to improve predictiveness of outcome. The SAT is a type of social test, and its purpose is to predict a certain social outcome -- student performance in college.
Not to say I did not do well on the last SAT exam I took (on the contrary, because the material is so basic), but why I had to spend the money and waste the time to do so in the first place just stinks.
The SAT costs less and is faster than testing via college credits. Would you test the inflation-level of tires by driving a car around and measuring its performance (efficiency, road-handling, straight-line traction, smoothness over bumps, road noise, tire-temperature-rise, etc.) because it is cheaper and easier than to "spend the money and waste the time" using a tire-pressure meter instead?
The article is spot on when it says that students are spending time gaming the SAT.
One way around that might be to give them working-memory and reaction-time/decision-time tests, instead.
http://www.google.com/search?q=%22reaction+time%22+jensen+%2...
http://www.google.com/search?q=%22working+memory%22+%22g+fac...
One of the purest tests of working memory is the Letter-Number Sequencing subtest of the Wechsler Adult Intelligence Scale (WAIS). It consists of simply reordering (in one's head) strings of random letters and numbers.
And yes, the SAT/ACT may still serve a purpose for many students, but I believe its importance is overemphasized.
Oddly enough, I would prefer something like WAIS to the SAT. Having taken it before, it's vastly more enjoyable (I also spent most of my k-12 education in a gifted school requiring a certain IQ, regardless of what that measures and how useful it is overall.)
Did you take the SAT after you had completed a ton of college classes?
[Edit: abridged]
Your rant reminds me of professional programmers who get insulted when asked to do fizz-buzz during an interview. What's the big deal? It's the quickest and easiest way to weed out the riff-raff. In my experience, the ones who get insulted are usually the ones who can't do it.
Those little tests don't account for documentation, architecture, scalability, readability, refactoring, etc. I don't give a damn if somebody can't come up with some little algorithm during the pressure of a test that could have easily just googled for, and also what is the best practices for that algorithm.
You are probably somebody new to the industry.
Of course a lot of poor coders can do fizz buzz.
But, if you can't code fizz buzz in under five minutes, I guarantee you are a bad coder.
It's a simple way to eliminate half your applicants. Sounds good to me.
Fizzbuzz and the SAT alike are to me just filling some very limited uses where they would come in handy - like if you ever DID ask fizzbuzz from an interviewee you'd at most be able to weed out people who don't know what the heck they're talking about, like some moron applying for a Java position that knows no Java. But to me the difference is that the SAT is like fizzbuzz applied to everyone regardless...if you had an interviewee that you knew in every way that fizzbuzz would be a waste of time (even that tiny amount of time) because they've proved themselves in some other way, you wouldn't ask. That's pretty much all I'm wanting. Some things are just not worth the effort because there's plenty else to see who they are and what they are capable of doing.
On the other hand, when I've already taken multiple semesters of advanced math classes and they want to know how well I score on a few dozen algebra questions, I find that slightly disturbing. If they wanted SAT Math 2 scores, I might understand. Otherwise it is pretty weird.
Also, a few years ago in high school my parents offered to pay for prep classes and I decided to because I didn't have much else to do during most of the summer. What was even more striking was that people NOT taking prep classes were in the minority - the only ones that didn't were the ones that were confident they could do it on their own. Even the poor ones managed to borrow a book or two to study.
A lot of stupid people can probably do well on SAT math because it's too easy. But I promise that it you can't score at least a 700 on it (i.e., demonstrate basic competence at elementary algebra) you are bad at math, no exceptions.
Exactly like a lot of stupid people can do fizz buzz, but if you can't do fizz buzz you are bad at programming.
But when you have years of high school or college, unless there's something huge missing, I don't know about how useful it would be to ask of it from everyone.
For the record, I almost fell asleep taking the SAT. It's been a while, but I think I scored a 790 on math. I vividly remember some of the people in the same room as me who were really worried they couldn't even break 5/600. Sooo...
Maybe they've fixed this bug, but here's an example. If you have a math problem like, (substituting x for actual words)
x x 5 x 5 x x x
And your choices are a) 5 b) 10 c) 15 d) 25
If this question is at the beginning of the test, its most likely the easiest solution, so pick 10. This is because the first questions are moved to the front because they are most easily solved by the worst test takers. These people, no matter how twisted their reasoning, most likely will come up with 10 by adding the two numbers they see, or just pick 5 because they just saw a 5
Towards the middle of the test you'd go with 25, because these are less likely, and necessarily wont be guessed by the dummies, but the people having reached this point will do something fancier like multiply the numbers together,
At the end of the test, where its hardest to reach, the answer would be 15 because you are least likely to come up with that answer.
This is the key to the whole hack- the answer to a difficult question cannot be accidentally guessed, otherwise, it would end up in the first questions, or just be thrown out. You see, they couldn't determine the level of difficulty of a question by reading it, they did statistically by who got it right, and how well that person did on the overall test, so if you had a question that was answered correctly by most takers who also got most of the other questions correct, and got wrong by those who didn't do well on the rest of the test, that was a "hard" question.
When this was uncovered the testing company freaked out and the move to computers couldn't come fast enough for them (they administer these tests on computers yet?) Its because the test can dynamically be calibrated for the test taker and you can't use the first questions, middle questions, last questions rule.
Grades? Those inflated things?
Recommendations? Wonder how long it will be before a family sues a teacher who gives a poor recommendation. HR managers already overwhelmingly provide positive feedback about former employees.
Work samples? Awesome, except that you could get someone else to do it for you.
Activities? Well, too many students are creating shell activities where they hold leadership positions. No meetings, no members to run against you for president, no hassle, no mess.
Not that the SAT is much better considering the test-prep that goes on and the fact that the College Board no longer indicates different test conditions.
Creating effective screens (both for school admissions and employers) is very difficult. You can create positive correlations between things - for example, those with higher SAT scores tend to be smarter. Doesn't mean that you don't have random idiots with high scores and smart people with low ones. But there is a positive correlation. As such, employers and schools will use that just as a life-insurance company would use the fact that you're a smoker to correlate with a greater chance of dying young. Some smokers live very long lives, but there is a positive correlation the other way.
And perhaps you're like me who didn't give a damn about what the school teaches, because other things seem more important, like entrepreneurship, and yet extracurriculars and character only make about 25-33% of the decision.
http://graphics8.nytimes.com/packages/pdf/education/20080922...
SAT Reasoning Test Critical Math Writing
Reading
Family Income
Less than $10,000 427 451 423
$10,000 - $20,000 453 472 446
$20,000 - $30,000 454 465 444
$30,000 - $40,000 476 485 466
$40,000 - $50,000 489 496 477
$50,000 - $60,000 497 504 486
$60,000 - $70,000 504 511 493
$70,000 - $80,000 508 516 498
$80,000 - $100,000 520 529 510
More than $100,000 544 556 537
Highest Level of Critical Math Writing
Parental Education Reading
No HS Diploma 421 445 418
HS Diploma 466 476 457
Associate Degree 484 492 473
Bachelor’s Degree 522 533 513
Graduate Degree 560 569 552
(College Board score report, 2007. Retrieved from http://www.collegeboard.com/prod_downloads/about/news_info/c...
report.pdf)http://www.nytimes.com/2008/09/22/education/22admissions.htm...
Commission Report .pdf:
http://graphics8.nytimes.com/packages/pdf/education/20080922...
NYT article comments:
http://community.nytimes.com/article/comments/2008/09/22/edu...
Excerpt from article:
Mr. Fitzsimmons’s group, which was convened by the National Association for College Admission Counseling, also expresses concerns “that test scores appear to calcify differences based on class, race/ethnicity and parental educational attainment.”
http://graphics8.nytimes.com/packages/pdf/education/20080922...
Average SAT Composite Score by Race/Ethnic Group, 2007
African American/Black 17.0
American Indian/Alaska Native 18.9
Caucasian American/White 22.1
Hispanic 18.7
Asian American/Pacific Islander 22.6
Other/No Response 21.6
(ACT National Score Report, 2007. Retrieved from http://www.act.org/news/data07/charts/index.html)