|
-Troika- posted:I think standardized tests filtering out people who can't read and do basic math is a good thing. If you cannot do those two things you should be in remedial classes at the local community college, not trying to get into somewhere where you will not be able to do the work and are thus wasting both your time and the university's. That'd be great if that's what they actually did.
|
# ? Sep 28, 2015 21:12 |
|
|
# ? Jun 8, 2024 08:28 |
|
Main Paineframe posted:Thanks for all this info! It's rare to get a glimpse into how this stuff works behind the scenes like that. What 'behind the scenes' information is there here? He also seems to have this weird misconception that his ability to game the test is somehow unique to him among high scorers. Basically all high scorers do this at some level, especially on the verbal section. And goddamn his stream-of-consciousness posting is insufferable. tehllama posted:I dunno, as someone who has always done extremely well on standardized tests the ability to see the question behind the question is definitely a huge part of it. Whether that's a reflection of better knowledge/ability with regards to actual content of the test, I personally don't really think so. I definitely know some very smart people who even at the graduate level just suck at certain types of questions and I have to think its because they read the question in a very different way, or fail to immediately see the examiners intent behind the question. I would actually argue that being able to correctly answer questions where you lack key pieces of knowledge is a greater sign of intelligence than merely a check on whether you know something or not. One is pattern matching, the other is just a knowledge check. Of course both knowledge and pattern matching are trainable skills to some extent but the latter maps more closely to what is considered 'intelligence' than the former. Lotka Volterra posted:That'd be great if that's what they actually did. The SAT I math section definitely provides a pretty good signal as to whether someone cannot do basic math.
|
# ? Sep 28, 2015 21:47 |
|
Cicero posted:As someone who has always been good at tests, I've long harbored a suspicion that "I'm not a good test taker" really meant "I'm not particularly smart/I don't actually study that hard", like, a good 60-70% of the time. It's much easier to admit being bad at tests than having a bad brain or being lazy. Are there any studies on this? -Troika- posted:I think standardized tests filtering out people who can't read and do basic math is a good thing. If you cannot do those two things you should be in remedial classes at the local community college, not trying to get into somewhere where you will not be able to do the work and are thus wasting both your time and the university's. All standardized tests do is create and enforce really bad ideas like those. Studying and test taking are both skills that can be learned and developed, and instead of actually teaching students how to master them, it's treated both by faculty and students as sink-or-swim. The goal of education should actually be, you know, education. Filtering students out from the "good" ones and the "bad" ones is completely antithetical to the goal of education in the first place. Remedial classes are just another brutal manifestation of the winners-vs-losers mentality of American education, where they do nothing except provide a place for students to languish further.
|
# ? Sep 28, 2015 21:50 |
|
The American education system is a joke anyway so who cares what measures they use for acceptance?
|
# ? Sep 28, 2015 21:56 |
|
Ideally a test should prove that you understand the concepts that you have been taught, but in reality this doesn't hold outside of college classes that don't ask multiple choice questions, ect. You have an entire segment of society that is good at taking standardized tests and could do good on the regardless of their understanding of a particular concept. How do you deliver tests that deliver and understanding of concept? Those are the only tests that should be offered.
|
# ? Sep 28, 2015 21:57 |
|
Chalets the Baka posted:Studying and test taking are both skills that can be learned and developed, and instead of actually teaching students how to master them, it's treated both by faculty and students as sink-or-swim. quote:The existing academic research base indicates that, on average, test preparation efforts yield a positive but small effect on standardized admission test scores. Contrary to the claims made by many test preparation providers of large increases of 100 points or more on the SAT, research suggests that average gains are more in the neighborhood of 30 points.
|
# ? Sep 28, 2015 22:05 |
|
There's a big difference between the type of family that sends their kids to SAT prep courses and the general population.
|
# ? Sep 28, 2015 22:11 |
|
Radbot posted:There's a big difference between the type of family that sends their kids to SAT prep courses and the general population. quote:The studies by Briggs, and Briggs and Domingue suggest that coaching is more effective for students with strong academic backgrounds and high socioeconomic status who underperformed on the PSAT
|
# ? Sep 28, 2015 22:23 |
|
Chalets the Baka posted:All standardized tests do is create and enforce really bad ideas like those. Radbot posted:There's a big difference between the type of family that sends their kids to SAT prep courses and the general population.
|
# ? Sep 28, 2015 22:25 |
|
silence_kit posted:I never took the SAT, but I took the ACT, and it wasn't some test of arcane Anglo culture like you all make it out to be. It had me read some articles and asked questions testing how well I understood the articles. The classic example is a series of questions involving a yacht. If you are a white New Englander, for example, even a poor one, you likely know the word yacht. If you are a black inner city kid, you may never have encountered the word at all. That wasn't what the question was supposed to be measuring, and the test makers likely didn't even notice it as a potential barrier. There are always implicit cultural biases in linguistic questions.
|
# ? Sep 28, 2015 22:26 |
|
blah_blah posted:I would actually argue that being able to correctly answer questions where you lack key pieces of knowledge is a greater sign of intelligence than merely a check on whether you know something or not. One is pattern matching, the other is just a knowledge check. Of course both knowledge and pattern matching are trainable skills to some extent but the latter maps more closely to what is considered 'intelligence' than the former. I guess it depends on what type of intelligence you want to assess. The SAT and ACT both kind of label themselves as assessing reading comprehension and high school level math ability, not intuitive intelligence. From a score reporting standpoint they're certainly not set up to do that well. Chalets the Baka posted:All standardized tests do is create and enforce really bad ideas like those. I think that quite a lot of the "test taking ability" is really a matter of intuition, and while certainly some basic skills (marking off answer choices you know are wrong, marking out extreme answer choices, etc) can be taught, it is pretty hard to teach students to intuit the exam writer's intent behind a question (which is probably the bulk of being able to answer a question where you lack key pieces of knowledge, as stated above). I totally agree that study skills can be taught and we probably don't do as good a job of it as we should but I very much doubt any amount of studying is going to bring the lowest performing students up to par with the highest performing students and you're still going to have students who need classes at a difference pace (faster and slower). Its not like many of the high performing students get there just because they have better study habits - many of them, especially at the high school level, have terrible study habits and get by even at the AP level just because content is delivered far slower than their maximum learning rate. tehllama fucked around with this message at 22:35 on Sep 28, 2015 |
# ? Sep 28, 2015 22:26 |
|
Trent posted:The classic example is a series of questions involving a yacht. If you are a white New Englander, for example, even a poor one, you likely know the word yacht. If you are a black inner city kid, you may never have encountered the word at all. That wasn't what the question was supposed to be measuring, and the test makers likely didn't even notice it as a potential barrier. There are always implicit cultural biases in linguistic questions.
|
# ? Sep 28, 2015 22:30 |
|
Strudel Man posted:I suspect you're thinking of the infamous runner:marathon::oarsman:regatta question.
|
# ? Sep 28, 2015 22:38 |
|
Cicero posted:To be fair, those sections are really the worst. Everyone hates them. The more esoteric vocab words in them don't test meaningful knowledge or meaningful intelligence.
|
# ? Sep 28, 2015 22:40 |
|
Cicero posted:To be fair, those sections are really the worst. Everyone hates them. The more esoteric vocab words in them don't test meaningful knowledge or meaningful intelligence. The ACT didn't even have a section like that which tested how well you can memorize vocabulary words. If I remember correctly, all those types of questions on the test allowed you to infer them from context.
|
# ? Sep 28, 2015 22:46 |
|
I think that the SAT is only really useful at identifying if there's something very wrong with a student*. Like, if a student gets a very low score it almost certainly means that they lack a bunch of basic skills, like arithmetic and reading comprehension. I think it loses its meaning at average and above-average scores, though. Using the old SAT as an example (with scores out of 1600, since that's the one I took), I don't think someone with a score of 1250 is necessarily going to do any better or worse in college than someone with a score of 1550. I think one fundamental problem is that the children of wealthier people (not necessarily wealthy, though; it's probably more useful to look at families that are just "financially stable"), due to the circumstances and opportunities of their upbringing, are likely to genuinely be more knowledgeable and capable than poorer children. If a student has a stable upbringing with more educational opportunities, of course they're usually going to end up being a better student. This is why things like affirmative action are necessary; if you don't forcibly give disadvantaged groups access to more resources and opportunities, they will always generally lag behind. Most students are not better students because they're just intrinsically better people or something; they're better students because they've had access to better opportunities or had less problems to deal with that would interfere with their education. The fact that some outliers exist doesn't change this (just to preempt some dumb "but I was poor and got good test scores!" replies). *While FilthIncarnate is correct that people would get low scores even if 100% of the test-taker population was college ready, in reality there actually are a large number of students who aren't college-ready, so this isn't as much of an issue. edit: Cicero posted:To be fair, those sections are really the worst. Everyone hates them. The more esoteric vocab words in them don't test meaningful knowledge or meaningful intelligence. The worst thing is that I think analogies are actually an incredibly useful test of intelligence, but only if you use vocabulary that all test-takers can be expected to understand. Interspersing esoteric vocabulary in an analogy section (which is how the tests actually were) is absurd and pointlessly conflating two different types of knowledge/ability in one type of question. While it's good that they removed them given the form in which they existed, I would rather they still exist and just use reasonable vocabulary. Ytlaya fucked around with this message at 22:57 on Sep 28, 2015 |
# ? Sep 28, 2015 22:54 |
|
I think they removed analogies in 2005.
|
# ? Sep 28, 2015 22:55 |
|
The problem for me with getting rid of test scores is that grades are so inflated in the vast majority of high schools that they're not very useful for most selective colleges once you filter out the obviously unqualified applicants. I think more essays are good, especially when you're able to interview all your applicants, but I don't think it's feasible everywhere. The article in the OP talks about looking at "community engagement and entrepreneurship," but that strikes me as even more class driven than test scores.
|
# ? Sep 28, 2015 23:24 |
|
Xandu posted:The problem for me with getting rid of test scores is that grades are so inflated in the vast majority of high schools that they're not very useful for most selective colleges once you filter out the obviously unqualified applicants. I think more essays are good, especially when you're able to interview all your applicants, but I don't think it's feasible everywhere. The article in the OP talks about looking at "community engagement and entrepreneurship," but that strikes me as even more class driven than test scores. Tests are a little bit gameable but they still seem less gameable than the alternatives.
|
# ? Sep 28, 2015 23:25 |
|
Yeah, I guess you can check it against letters of recommendation to see if writing level/content is plausible, but there's no doubt that essay cheating is rampant. edit: SAT prep gained me like a 150 points, though some of that might have just been from having more familiarity with the test. Xandu fucked around with this message at 23:29 on Sep 28, 2015 |
# ? Sep 28, 2015 23:27 |
|
Cicero posted:I mean, short of thoroughly interviewing the applicant (do undergrad admissions programs do this?) Ivy Leagues do, I remember a friend had an interview with someone from Columbia.
|
# ? Sep 28, 2015 23:52 |
|
Cicero posted:As someone who has always been good at tests, I've long harbored a suspicion that "I'm not a good test taker" really meant "I'm not particularly smart/I don't actually study that hard", like, a good 60-70% of the time. It's much easier to admit being bad at tests than having a bad brain or being lazy. Are there any studies on this? . . . What could such studies possibly study? How would you go about making a measurable, quantifiable comparison between 'real' intelligence and test results for a thousand students?
|
# ? Sep 29, 2015 00:02 |
|
The three or four interviews I did for college involved me driving from a relatively well-off area in one of the top 25 richest counties to an extremely well-off area to sit in giant mansions and listen to extremely white people tell me how their sons and/or daughters loved the school. I'm white as hell and it was still really uncomfortable. Interviews are dumb.
|
# ? Sep 29, 2015 00:06 |
|
computer parts posted:Ivy Leagues do, I remember a friend had an interview with someone from Columbia. I was told at my undergrad interviews for Ivy's that the interviews have basically no weight whatsoever in the admission process. They are done by local alumni and not official admissions people so there may actually be some truth to that. Almost everyone who interviewed me was an rear end in a top hat in some way. I interviewed at Cambridge (on campus) and it was extremely different. There were 3 interviews if I remember correctly, and I think in only one did I get asked a "why your major, why Cambridge?" The rest (since I was interviewing for engineering) were just "here is a problem, explain how to solve it." There was one about finding a dimension using only certain geometric shapes, one about airflow in a room under various conditions, and a few others that I don't remember. The only US non Ivy's I interviewed at were for scholarships. It just isn't really a thing at most US undergrad institutions, probably mostly because of applicant numbers. The big state schools may accept 10,000+ students, and they're narrowing from 20,000+ applicants. e: KaptainKrunk posted:The three or four interviews I did for college involved me driving from a relatively well-off area in one of the top 25 richest counties to an extremely well-off area to sit in giant mansions and listen to extremely white people tell me how their sons and/or daughters loved the school. I'm white as hell and it was still really uncomfortable. Holy poo poo this was basically exactly my experience.
|
# ? Sep 29, 2015 00:13 |
Phyzzle posted:. . . What could such studies possibly study? How would you go about making a measurable, quantifiable comparison between 'real' intelligence and test results for a thousand students? There are a variety of statistical and methodological tests one can run to evaluate different forms of measurement validity. While directly proving nomological/test validity is impossible, you can prove that a measure is more or less consistent or unbiased in its measurement, and correlate it with other measures that you believe measure correlated concepts/constructs. Discendo Vox fucked around with this message at 00:44 on Sep 29, 2015 |
|
# ? Sep 29, 2015 00:41 |
|
Cicero posted:It's not "good vs bad", it's "more knowledgeable vs less knowledgeable" or "more intelligent vs less intelligent". Believe it or not, those are meaningful things. Refusing to do any kind of filtering and just dumping everyone into the same bucket means you'll just teach to the lowest common denominator. If you believe that people are innately born with what you believe to be intelligence, and those who do well on standardized tests are "intelligent" and those who don't are "not intelligent", then all you're doing is creating artificial academic strata. You've already decided by high school who the winners and who the losers in life are; those who get further opportunities to learn and those who are shut out of them. Treating standardized testing as a zero-sum game is asinine. Naturally those who excelled at their testing, got their opportunities handed to them on a silver platter, and deign to entertain the idea that anyone less than them could possibly perform on their level also espouse how instrumental that environment is in determining intelligence.
|
# ? Sep 29, 2015 00:44 |
|
I like tests because test taking is an art, and it is one I am great at. Any test that just has you fill in bubbles isn't a judgement of your level of knowledge, it is just testing your ability to understand the test.
|
# ? Sep 29, 2015 01:58 |
|
What is this crap, I'm not voting in a one choice poll.
|
# ? Sep 29, 2015 02:07 |
|
Chalets the Baka posted:If you believe that people are innately born with what you believe to be intelligence, and those who do well on standardized tests are "intelligent" and those who don't are "not intelligent", then all you're doing is creating artificial academic strata. You've already decided by high school who the winners and who the losers in life are; those who get further opportunities to learn and those who are shut out of them. If you don't do well on the SAT/ACT, you are free to go to community college and should be able to effortlessly transfer into your state flagship if you are actually smart. Nobody is locked out of anything if they are willing and able to do the work, and these people will save a poo poo ton of money on college to boot.
|
# ? Sep 29, 2015 03:01 |
|
One thing some of you might find interesting is the way that standardized testing is approached in other countries. The way the US does it is actually particularly bad, since the SATs don't have much relation to coursework material that students are learning throughout their last few years of high school, and it is the test itself that is of such extreme importance. So you have a 4 hour test that can make or break college applications, and a students GPA (which should represent their performance over a much longer period of time; and should therefore be a much more reliable estimate of drive and ability etc) which is in fact fairly meaningless since there is no mechanism to compare across schools. The way it is done in Australia, for example, is that there are end of year examinations on the courses students took during 11th and 12th grades, with the questions actually focusing on the material in those courses (rather than random SAT pseudo IQ-test malarkey). So it's sort of like the IB exams, but every student is taking them (rather than the people who happen to be priviledged enough to be in AP/IB classes). The results of these exams is then used to normalize school GPAs, and then half weight is given to the exams, half to the school performance / GPA. Definitely has it's weaknesses (there is an extreme advantage in going to a top private school or selective government school rather than being one of a handful of high achievers at a random public school, since in the latter case how well GPA's normalize for a given school can literally come down to the exam performance of 1 student, the valedictorian, if they happen to be significantly more capable than their nearest competitor), but it's way better than the crapshoot that's US undergraduate admissions. Even grad school in the US has some laughable standardized testing requirements. It's honestly embarrassing to everyone involved that the GRE is even a thing.
|
# ? Sep 29, 2015 03:55 |
|
Cicero posted:Coaching/cheating for essays seems like about a thousand times easier to do than for testing, though. I mean, short of thoroughly interviewing the applicant (do undergrad admissions programs do this?) and having them write an essay on the spot, how can you tell if someone else wrote the essay for them? Our standardized diploma exams in Alberta involved either essay writing (for English and history) or multi-part written responses (for math and science). The need for standardization meant that it was still reasonably simple to game the exam, but giving people three hours to write an essay on the spot is by no means infeasible. Hell, I did it a few times in university too. Even multiple choice tests aren't inherently lovely. One of the most challenging tests I ever wrote was multiple choice, to get my pilot's license. It was carefully crafted to make sure that any of the "common shortcuts" would bite you in the rear end. You could prepare for it, but the only way to do so successfully was to practice the skills they were actually testing. So, what's the difference? Well, in school, everyone has a vested interest in making sure you do alright, but for a PPL, they want to make sure you aren't going to kill yourself or anyone else, and they don't give a fat gently caress if you fail. Being honest, a difficult MC test is far more daunting than an essay; you don't get gently caress all for being 90% correct. Actually challenging students, and making it difficult for students of privilege to game the test in some way, is not advantageous to anyone involved, so it's not done. No one wants to explore the factors that, far before a student enters college, affects their education. That doesn't mean it's impossible to do, just that it will never change unless the influences on the system can be altered.
|
# ? Sep 29, 2015 04:02 |
|
PT6A posted:Our standardized diploma exams in Alberta involved either essay writing (for English and history) or multi-part written responses (for math and science). The need for standardization meant that it was still reasonably simple to game the exam, but giving people three hours to write an essay on the spot is by no means infeasible. Hell, I did it a few times in university too.
|
# ? Sep 29, 2015 04:04 |
|
Shakugan posted:Even grad school in the US has some laughable standardized testing requirements. It's honestly embarrassing to everyone involved that the GRE is even a thing. In fact it's gotten to the point where over the years most of my professors admitted that they pretty much ignore the GRE. The tests we currently take aren't evaluating much useful info in the eyes of the educators themselves.
|
# ? Sep 29, 2015 04:04 |
|
Cicero posted:Xandu was talking about application essays, not exam essays. I did two of those in a timed "exam" environment too. This isn't an insurmountable problem, it's just that the powers that be have absolutely no interest in doing things the right way.
|
# ? Sep 29, 2015 04:11 |
|
Xandu posted:The article in the OP talks about looking at "community engagement and entrepreneurship," but that strikes me as even more class driven than test scores.
|
# ? Sep 29, 2015 04:22 |
|
tehllama posted:I think they removed analogies in 2005. pretty sure i did them in 2007 also as far as the essay part goes i bullshitted out like eight sentences about sandwiches and got a 620, poo poo is dumb and bad
|
# ? Sep 29, 2015 04:22 |
|
PT6A posted:I did two of those in a timed "exam" environment too. This isn't an insurmountable problem, it's just that the powers that be have absolutely no interest in doing things the right way.
|
# ? Sep 29, 2015 04:37 |
|
OneEightHundred posted:I'm not sure how you can bypass the "class-driven" problem in the first place as long as class disparities affect college preparedness, especially via badly-underfunded school districts. Why focus on admissions standards when there are blatant problems at the primary/secondary level like, say, funding public schools with local tax revenues, de-facto segregation via district line fuckery, and the public school system in the US being pretty crap in general? This is the real issue: fixing the class and racial disparities at the college admissions level is just putting lipstick on a pig. The problem is that we're failing students at a much earlier, much more basic level in most cases. I don't think it's a problem the education system can fix on its own, either: we're talking about a need for serious investment to prevent childhood poverty, and the need to show students, early on, that they can succeed in life through education. By the time students get to high school, can you really fix the fact that they've been systematically hosed over for the better portion of their lives, no matter what you do to the admissions process?
|
# ? Sep 29, 2015 04:40 |
|
Shakugan posted:One thing some of you might find interesting is the way that standardized testing is approached in other countries. The way the US does it is actually particularly bad, since the SATs don't have much relation to coursework material that students are learning throughout their last few years of high school, and it is the test itself that is of such extreme importance. The US has a huge population divided across states with a substantial degree of educational autonomy, and gigantic variance in school quality at both the secondary and postsecondary level. It's a lot easier to do things in Australia (or Canada) where almost all decent universities are public and most would rank somewhere in the 70-90th percentile of universities in the US. Even in Canada, nation-wide comprehensive examinations aren't in place (even provincial-level testing is largely disappearing) and the net effect is favorable to students at top public or private schools which can freely grade inflate (i.e. it's a long ways from being a reliable estimate of 'drive and ability', and in particular the curriculum is too simple to differentiate between top students, which is why AP/IB courses have also become popular here).
|
# ? Sep 29, 2015 04:48 |
|
|
# ? Jun 8, 2024 08:28 |
|
OneEightHundred posted:I'm not sure how you can bypass the "class-driven" problem in the first place as long as class disparities affect college preparedness, especially via badly-underfunded school districts. Why focus on admissions standards when there are blatant problems at the primary/secondary level like, say, funding public schools with local tax revenues, de-facto segregation via district line fuckery, and the public school system in the US being pretty crap in general? That's a good point and I sort of thought of it when writing my post, but I guess because that's clearly the aim of the college here. They're are explicitly trying to correct for that, so the question becomes, what's the best way to do that? In this case, dropping the SAT requirement increased persons of color and first generation students, but it's also a very small, selective liberal arts college, so it's not clear how relevant it is for other schools. And to be honest, it goes far beyond schools. Lots of evidence that teacher/school quality is fairly marginal in outcomes outside of the extremes and thedifferences go back really to economic disparity.
|
# ? Sep 29, 2015 04:52 |