Monday, September 25
Monday, September 25, 2006 | 
Posted by
Anonymous
I have long been a protestor of standardized tests -- even when I took my first PSAT when strawberries was a mere small red dot on the vine. Not because I have never been one of those "I never study and get a 1600" types, but because I do not think they actually measure ability to do well.
Take the GRE for grad school. Now, in my discipline, there are never multiple choice situations. And the quantitative aspects are statistics, complicated statistics that are handled by stat programs. Sure basic math skills are important, but when ever has an academic really figured out an important geometry problem, and then, without a calculator.
Ok, so the verbal section. All important, sure, in theory. But when one is entering a grad program, they should have a) done well in undergrad, b) have demonstrated a capacity to conduct effective research, and c) ability to write well.
The reading comprehension asks questions about what is the next logical paragraph, or title. Then in those lovely books prepared by ETS, they explain that only Y can come after X. I disagree. First of all, writing has a style. An argument can be explained effectively in more that one way. Perhaps Z after X goes better together, even though chronologically Y comes first. This is all a question of style, journal preferences, and topic. And don't get me started on the Writing Analysis section....
And then the ridiculous emphasis that NSF and other organizations place on these scores -- NSF Doctoral Grants /Research grants should not take into strong, if any, consideration standardized tests, but research capability, etc. Maybe the SAT is needed to generalize across different school systems and their grading, and since students still do not know what direction their studies/life will take, general capability is good. But in a specialized field?? I back the idea of subject tests, but carried out in a different manner.
With all the test prep books, the test prep classes, and the test prep costs, it seems to me that aside from being almost useless in determining success in graduate school, it is just one big money making scheme.
Take the GRE for grad school. Now, in my discipline, there are never multiple choice situations. And the quantitative aspects are statistics, complicated statistics that are handled by stat programs. Sure basic math skills are important, but when ever has an academic really figured out an important geometry problem, and then, without a calculator.
Ok, so the verbal section. All important, sure, in theory. But when one is entering a grad program, they should have a) done well in undergrad, b) have demonstrated a capacity to conduct effective research, and c) ability to write well.
The reading comprehension asks questions about what is the next logical paragraph, or title. Then in those lovely books prepared by ETS, they explain that only Y can come after X. I disagree. First of all, writing has a style. An argument can be explained effectively in more that one way. Perhaps Z after X goes better together, even though chronologically Y comes first. This is all a question of style, journal preferences, and topic. And don't get me started on the Writing Analysis section....
And then the ridiculous emphasis that NSF and other organizations place on these scores -- NSF Doctoral Grants /Research grants should not take into strong, if any, consideration standardized tests, but research capability, etc. Maybe the SAT is needed to generalize across different school systems and their grading, and since students still do not know what direction their studies/life will take, general capability is good. But in a specialized field?? I back the idea of subject tests, but carried out in a different manner.
With all the test prep books, the test prep classes, and the test prep costs, it seems to me that aside from being almost useless in determining success in graduate school, it is just one big money making scheme.
8 comments:
I completely agree. Those exams are a more reliable measure of class and status than actual ability.
I remember transferring schools, a month after a few short stories of mine got published. When I took the all-essay English competency test, I failed.
Now, I'm not smug enough to think that publishing a couple dippy short stories in a couple dippy journals makes me the king of all language. But you'd think I would have at least passed out of remedial English at that point. But no, according to this school, I had to retake high school English. It's been a billion years since that, and I'm still puzzled.
Oh yeah, the ETS is one big evil force. They are completely ridiculous. And yes, Apri put it well, it's a much better measure of class and status than anything else. Oh, and national origin, since no other country I know of tests students in this way so you're adding a significant handicap to foreigners' changes of doing even remotely well.
So yes, it's a BIG money-making machine, completely disgusting and should be taken less and less seriously as far as I'm concerned.
I'm a foreigner (being English), and I found the verbal section of the GRE had questions that didn't make sense to me, for purely cultural reasons. It wasn't quite on the level of the sub-$1000 questions on Who Wants to be a Milliner, which basically test whether you grew up in the US, but it wasn't far off.
That said, my personal experience makes me worry about this (from the original post):
"when one is entering a grad program, they should have a) done well in undergrad, b) have demonstrated a capacity to conduct effective research, and c) ability to write well."
I got taken on despite (a) being untestable (there being no clear equivalence in undergrad grading norms between the US and UK) and (b) being irrelevant to my undergrad (though not the grad course I was applying to). The only thing I could fall back on, really, was writing well, and the only evidence for this was my Statement of Purpose and my GRE score. I fear that using the criteria mentioned here I might have been worse off, as a foreigner, than under the current scheme.
The GRE's are pretty bad, I agree. Two students from my department both applied for NSF grants. The one with really good GRE scores got it, though he was not a promising candidate, and not great at fieldwork. The other was amazing at theory and fieldwork, but bad at tests. Valuing them too highly is definitely a risk.
As noted above, I'm definitely against these standardized tests. However, Strawberries made a point I want to follow-up on. Recommendation letters are also somewhat cultural, if you can believe it. Or they reflect cultural norms. So in certain other countries (probably most), it is much less likely that teachers will praise the way they do in the US. They would never say someone's in the top 5-10% even if they were. So that leaves the candidate in a difficult position.
I'm not sure how one can deal with all this, it gets pretty complicated.
Thanks for the clarification, strawberries and turquoise stuff. I was actually careful to make sure that I had one letter of recommendation by an American, more because I suspected that there might be a standard form for such things than because I considered (at the time) that norms of praise would be different. Now I know better, I'm even happier I thought of it.
The lack of commensurability between US and UK grading goes beyond different numbering schemes: when I did my undergrad degree, you took a degree in one subject and were given a single result based on your performance in that subject - there was no equivalent to the course-load, GPA or what have you. It's such a fundamental difference that it can be hard to wrap one's mind around it. No majors, no minors, just the subject chosen at the time of application to university. Further, the possible grades were pass, fail or distinction; not a very clear metric in numerical terms. Perhaps strangest of all, I later found out that the professors who approved my application here in the US never tried to draw a clear parallel - they didn't know abot any of these differences...
Post a Comment