Fundamental flaw in GRE reading comprehension test

[Helpplease123]

Ok, so what I'm hearing is that the passages should NOT clearly communicate, but deliberately obfuscate (eg by removal of headlines, usage of long-winded sentences etc)?

Correct. It is a method of ranking candidates abilities in comprehension when things haven't been spelled out to them. If the passages were too clear and straightforward everyone would do just fine, it would be very self-esteem boosting but frankly why bother having the test.

[bigant]

Not necessarily deliberately obfuscate to all, but the passages are NOT there to communicate.

Correct. It is a method of ranking candidates abilities in comprehension when things haven't been spelled out to them.

It would be great if we could get someone from ETS to confirm that it is NOT the aim of the passages to communicate.

Anybody know anybody at ETS?

Edited December 22, 2011 by bigant

[Kitkat]

With due respect to KitKat, the purpose of a headline is not ONLY to help you choose which article to read. Even if you are reading a single article, a headline is still useful and necessary.

Would you be able to give me an example of how a headline would help you find the details in a passage that you need? I am not here to deny the fact that that the headline is a useful and necessary tool in the real world. But how is it there beyond the point of knowing which article to read and not read. It is a simple line at most. There is a limited amount of space for a title, and therefore in my mind, a very limited and very general overview that a title can give you. If you can give me an example of how, in the example of something like the GRE passages, a title can help you, then maybe I can understand the point you are making better.

My point is still this, that if you need a title to understand a passage then maybe you don't belong in grad school. While other arguments about unclear sentences and the like have merits, I think the arguments for headlines at this point are still weak. But at the same time, poor sentence structure and bad punctuation might still be interpreted correctly. I just feel that a dependence on headlines shows that you might not have the reading skills necessary for graduate school.

[finknottle]

My point is still this, that if you need a title to understand a passage then maybe you don't belong in grad school.

Wow! This thread is still active.

Looks like it's 'Make your generalization' day. I don't think anyone is suggesting that one 'needs' a title to understand the passage or that there is a 'dependency'. It's just a matter of convenience. That's how most people process information. They scan the title, look for a few key words and form a skeleton of the passage in their head. Now, if you take away the title and obfuscate the language, it doesn't make the problem more difficult or challenging in a true sense. It just makes it more inconvenient and tedious. Some people can adapt to it quickly and some might take more time. In any case, it has got little to do with 'belonging' to grad school. If I am asked to take a piss in the dark, and if I miss my aim and end up splashing it all over the place, would I be deemed unfit for civilized society? The only thing I can agree with is that it is an arbitrary standardized test that puts a number against an applicant. Besides, grad school is a pretty generic term. It can mean a one year program, or a PhD that can take half a decade or more. The emphasis of the program can also vary widely. Hence, parsing a large amount of verbiage might not be an integral part of all programs.

[bigant]

Now, if you take away the title and obfuscate the language, it doesn't make the problem more difficult or challenging in a true sense. It just makes it more inconvenient and tedious.

If I am asked to take a piss in the dark, and if I miss my aim and end up splashing it all over the place, would I be deemed unfit for civilized society?

Perfectly put, finknottle!

Edited January 2, 2012 by bigant

[emilyrobot]

The thing is, the GRE isn't supposed to test your reading comprehension, not really. It's supposed to predict your ability to do well in grad school, and depending on whose research you read, it does an Ok-to-terrible job at this. From my (very brief) review of the research, it seems to be so that there is certainly a difference in, say, the performance of folks who scored in the 25th or below and those who scored in the 75th percentile or above, but I don't think even ETS claims that the test is useful for predicting differences between, say, the 89th and 92nd percentile scorers. The subject tests are somewhat better than the quant/verbal tests, and the analytical section was pretty useless (which is why they dropped it).

If there was research that demonstrated that people who can pee without mess in the dark consistently earn high GPA's in graduate schools, adcoms might call you in for a pee test. If there was a similarly strong correlation between people who could accurately answer questions about reading passages where every 4th word was written in Arabic, and people who could successfully complete grad programs, then maybe that's what the test would look like. Neither test would be fair, neither test would contain good writing, and neither would say much about your reading comprehension per se, but if the correlation was really high, grad schools would be interested in it.

[ANDS!]

That "article" was ridiculous, and no doubt written by someone frustrated with the overall ridiculousness of the "old" Verbal GRE section. Make no mistake, I am no fan of the GRE's, but as has been pointed out, the dense writting on display isn't some conjuration of literary sadists, but the real way in which information is presented in scholarly papers; that's just the way it is. Putting a "title" on the paragraph defeats the purpose; you are MEANT to suss the meaning out of the small passage, and no you aren't meant to read it "just once". That is ludicrous. No one expects you to read a passage "just once" in graduate school, why would you think that is so here?

Thankfully the revised GRE has eased up on the overly long reading comprehension sections (in the three verbal sections, I came across maybe two long passages that had more than three questions attached to it); the passages are much shorter (often one or two paragraphs) and sometimes only attached to a single question. I still think standardized tests are a poor means to set baseline measures of acceptance, but at the least - it (being the GREs) has gotten better.

[bigant]

The thing is, the GRE isn't supposed to test your reading comprehension, not really.

Why call it a reading comprehension test then?

I agree with your point about correlation, but then ETS would have to admit that it's not really a reading comprehension test.

Edited January 5, 2012 by bigant

[Eigen]

Because it's a test based around the skills commonly referred to as reading comprehension? Technically, the test is called "Verbal Reasoning", with one of the skillsets it claims to test being reading comprehension.

[emilyrobot]

Right, it's not actually called a reading comprehension test. It's a graduate admissions test, designed (like IQ tests and SAT/ACT tests) to assess your aptitude, not the skills you've already acquired.

I don't think too highly of the GREs, honestly. It's certainly not perfect test by a long shot. Some of the validity research is kinda self-fulfilling (for instance, if the receipt of fellowships is dependent in part on GRE scores, it makes sense that people with higher GRE scores would be more likely to finish their degrees, because they're more likely to get funding). A lot of the weight that admissions programs put on GRE scores has less to do with people thinking that the GRE is a great test, and more to do with how heavily US News and World Report weights average GRE scores of students when ranking programs. So, I'm not super invested in defending the test itself.

However, as an applicant to PhD programs in education the following is kind of a hobbyhorse of mine: these tests are extensively researched, by both ETS and independent evaluators. That's the reason everybody has to sit for an experimental section, for example. Populations and averages are compared, the test is normed, individual results are compared to future performance and results on other, similar tests. That doesn't mean they're infallible, of course--some of the research will even demonstrate specific flaws in the test. None of the questions are just made up by some dude, and it seems pretty arrogant to me when people critique standardized tests based on what makes sense to them. Not that you can't criticize! But you should have a basic understanding of the ways that tests are designed and written and some ability to read the research that looks into their effectiveness. I know hard science types don't always take the social sciences terribly seriously, but we do have methods, some of them rigorous, to look into these types of problems. I try not to bother Biologists with my kooky theories about how to revise taxonomies of living things (let's categorize by size and color! That just makes sense to me!), and I wish my own discipline ever got the same measure of respect.

Whew, I guess according to the rant above, I have Feelings About This. To be clear, the above rant isn't directed to any individual person here, please don't take offense!

[bigant]

these tests are extensively researched, by both ETS and independent evaluators. That's the reason everybody has to sit for an experimental section, for example. Populations and averages are compared, the test is normed, individual results are compared to future performance and results on other, similar tests. That doesn't mean they're infallible, of course--some of the research will even demonstrate specific flaws in the test. None of the questions are just made up by some dude, and it seems pretty arrogant to me when people critique standardized tests based on what makes sense to them. Not that you can't criticize! But you should have a basic understanding of the ways that tests are designed and written and some ability to read the research that looks into their effectiveness. I know hard science types don't always take the social sciences terribly seriously, but we do have methods, some of them rigorous, to look into these types of problems.

Take one item for now: passage titles. Presumably, ETS's experimental tests have not included passage titles. Is there testing research that proves that passage titles negatively co-relate with graduate study success?

[Kitkat]

Take one item for now: passage titles. Presumably, ETS's experimental tests have not included passage titles. Is there testing research that proves that passage titles negatively co-relate with graduate study success?

bigant, you seem very focused on the idea of titles. While I agree that in real life titles are helpful, I have yet to see you make a good argument for having them on the test. While the test is not the best, people agree with you on that aspect, I do not see how that has anything to do with titles. You seem to argue against the idea of not having them, you don't seem to have a good explination of how they are useful in this testing environment.

They want to you to read the passage and do something like give them a title (aka give a one sentence summary). Giving you a title would defeat the point of this. They want you to read the passage and do something similar to 'What kind of title would you give this' but with more complicated questions.

Just tell be with examples, why are titles important for this type of test?

[brick]

As one who writes for a living, I have to agree that GRE passages are convoluted or what some refer to as "dense." In almost all cases, they lack clarity on first blush; however, a test - by its very nature - must be this way. The measure of a test's success, though, is ensuring passages have only one logical interpretation. In my limited experience, the GRE succeeds in this regard; the underlying reasoning the passages represent is not convoluted, is quite simple, and can generally be interpreted correctly in only one way. It is the way in which the underlying reasoning is expressed in the writing that *is* (intentionally) more difficult to grasp.

As for the oft-cited example in this thread...

"The increase in the numbers of married women employed outside the home in the twentieth century had less to do with the mechanization of housework and an increase in leisure time for these women than it did with their own economic necessity and with high marriage rates that shrank the available pool of single women workers, previously, in many cases, the only women employers would hire."

...the argument is simple, but it is presented in a manner more difficult to comprehend than is necessary. (The portions I struckthrough could be eliminated without losing any meaning while also decreasing the number of words to stumble over.) Instead of making it easier to understand, I will make it even more complicated:

"The lack of a decrease, and even the lack of a maintenance of a steady state, in the numbers of non-single women employed outside the home in the twentieth century had less to do with the proliferation of electrical appliances throughout homes in the Western world, a decrease in the amount of time required to do household work -- which typically belonged to the distaff's side -- and an increase in leisure time than it did with their perception of economic need and with the fact that fewer women were remaining single, a phenomenon that shrank the availability of employers' previously-relied-upon pool of single women workers, which were often the only women those employers would hire."

The underlying idea is the same, but the passage is even more verbose and (a bit) more convoluted than the original. The point is that the idea is obscured by a rambling string of words and backwardly-worded phrases like "lack of a decrease", "non-single women", and "fewer women were remaining single". The GRE tests a person's ability to comprehend stuff that's not easily comprehensible.

...you could make the same argument about "unfairness" for the GRE's choice of vocabular in the 'fill in the blank' portions of the test: if the test's aim is to measure verbal reasoning, why provide answer choices comprising words you will NEVER see or use again in your life? Well, because most people will get the right answers if simple vocab is used. The reasoning aspect of the questions isn't difficult -- the vocab is. Thus, the GRE also tests vocabulary.

All in all, if you can't get over all this and just suck it up, you're going to suck at life. Unless you start your own commune.

[asleepawake]

Whoa! This is an intense thread.

I personally love the reading comprehension questions. They are, for me, the easiest part of the entire GRE experience -they're almost fun, and I can completely understand their usefulness. I do not like text completion questions (when will I ever need to find the missing word?), math is... math, and I feel rushed during the AW. It may be that I come from an English department, but I cannot honestly say that any of the passages I have read on the GRE or any practice test have been have been anywhere near as difficult as some of the readings I have done in my MA program. The GRE passages are often short and light, if not especially straightforward, and easily understood by a strong reader regardless of disciple.

I cannot imagine how "headlines" would help with anybody. In fact, having the title of a work may simply waste your time. You are reading a paragraph or two. The title will not refer only to this small section but to the larger work from which it came, and for this reason I agree with ETS's choice to simply not include such titles.

The link posted in the start of this thread is ridiculous, and the author's argument is pretty much that ETS should make reading passages easier to comprehend on a test that is specifically designed to test a reader's ability to understand meaning. In all of the passages that ETS selects, the meaning is there. It is sitting right on the computer screen in front of you. You have to sort out this meaning, and you can do that even if you have no idea what a spinning jenny is.

In the specific case of the spinning jenny, it is painfully obvious in the reading that it is a piece of technology that made domestic work easier or faster. You don't need to picture the exact machine.

Edited June 11, 2012 by asleepawake

[bigant]

In almost all cases, they lack clarity on first blush; however, a test - by its very nature - must be this way.

The Introduction to the Analytical Writing Section of the GRE® revised General Test (Page 8, The form of your response, para 2, last sentence) says :

"What matters is not the number of examples, the number of paragraphs, or the form your argument takes but, rather, the cogency of your ideas about the issue and the clarity and skill with which you communicate those ideas to academic readers."

Clarity matters if it's the candidate doing the writing. But if it's GRE passage writing, clarity does not matter? A case of double standards.

In all of the passages that ETS selects, the meaning is there. It is sitting right on the computer screen in front of you.

Here's brick's modified "sentence" again:

"The lack of a decrease, and even the lack of a maintenance of a steady state, in the numbers of non-single women employed outside the home in the twentieth century had less to do with the proliferation of electrical appliances throughout homes in the Western world, a decrease in the amount of time required to do household work -- which typically belonged to the distaff's side -- and an increase in leisure time than it did with their perception of economic need and with the fact that fewer women were remaining single, a phenomenon that shrank the availability of employers' previously-relied-upon pool of single women workers, which were often the only women those employers would hire."

Since "the meaning is there", I'm sure the highly educated examiners could figure out the meaning, had I written such a sentence in a test. What are the chances that I'd get a good score?

[asleepawake]

I find the comparisons between the AW writing of the test and the reading comprehension sections to be illogical. They are testing two very different skills, and they are judged accordingly. I don't think the AW test is especially useful, and a lot of schools seem to agree, as they often ignore it or have no cut-off as long as you don't score very, very low.

As another user pointed out, you're not reading imaginary texts by GRE text writers sitting around twirling their mustaches. These are excerpts from real academic journals. This is the kind of material you will read all the time in graduate school. You need to be able to comprehend it quickly.

The more needlessly complex sentence that brick has written is absurd - sure, we can keep making this material more and more needlessly complex. But we're talking about the material on the test, not the most contrived sentences that could possibly exist. I disagree that any of the material that actually appears on the test is, however, so needlessly complex as others are suggesting. The specific quote about women in the workplace is not a perfect sentence, but it's not bad. I think it would be OK on the AW. The end gets a bit wonky, but I think it's pretty clear. It is something I would realistically read in and not think twice about in my studies. If it's not clear on the first reading, you read it again. I usually tackle the RC questions by reading the question, then the passage, then the question again. Finding the correct answer often involves some looking back and forth and I usually end up rereading parts of the passage to instill the confidence I need to move to the next question. Sometimes my first instinct is wrong, but I usually catch this by returning to the passage.

I'm surprised to hear so much criticism directed at the passages but almost none at the questions. The questions are not something you will see in grad school, but you certainly will be expected to rephrase and infer things about readings that are complex and perhaps even fatally flawed. The questions attempt to replicate that need, but they're certainly not perfect. As brick points out, the gre is generally successful in providing texts that are not ambiguous in meaning. As I said, the meaning is there.

It seems like what we're really having here, when the AW section is brought into the question, is a debate about scholarly writing in general. Should we aim to write more like Orwell or Derrida? Regardless of your answer, we have to learn to read all of these things if we plan to succeed in graduate school.

If the clarity were increased in the RC passages, what would the result be? Test takers might be a bit less stressed taking the test, but they are still going to receive a percentile score. The old quantitative test had this exact problem - some 4-6% of test takers scored 800, so the highest percentile was 94%, if I remember correctly. The test was too easy, and it has been adjusted for the new version. This problem would return with the RC/Verbal section if the passages were too easy for the average test-taker to comprehend. If the passages were rewritten to be completely clear to all readers immediately, they would just have to ask harder questions. But if the passage is easy, how can you ask a more difficult question about its interpretation? There is nothing to untangle, nothing to comprehend.

Edited June 13, 2012 by asleepawake

[bigant]

The more needlessly complex sentence that brick has written is absurd - sure, we can keep making this material more and more needlessly complex.

Brick has shown that the original sentence is also needlessly complex, by editing it:

"The increase in the numbers of married women employed outside the home in the twentieth century had less to do with the mechanization of housework and an increase in leisure time for these women than it did with their owneconomic necessity and with high marriage rates that shrank the available pool of single women workers, previously, in many cases, the only women employers would hire."

It's a matter of opinion whether or not a passage is unnecessarily complex. ETS needs to lay down some baseline standards for scholarly writing used in passages, because many "scholarly" writers are not trained in writing (as opposed to being literate) and write however they want, without any regard to clarity.

[BrianM]

ETS doesn't care about anyones opinions. Only people's opinions that will cause them to make more money. If they could somehow weasel a harder section into the test they would, as long as it caused people to spend money on classes and tests.

Its really sad this test even exists, I don't believe in the whole, "it differentiates between 2 similar candidates". To me its all about money and I am sure the universities get something out of this for making students take it.

[asleepawake]

Brick has shown that the original sentence is also needlessly complex, by editing it:

"The increase in the numbers of married women employed outside the home in the twentieth century had less to do with the mechanization of housework and an increase in leisure time for these women than it did with their owneconomic necessity and with high marriage rates that shrank the available pool of single women workers, previously, in many cases, the only women employers would hire."

It's a matter of opinion whether or not a passage is unnecessarily complex. ETS needs to lay down some baseline standards for scholarly writing used in passages, because many "scholarly" writers are not trained in writing (as opposed to being literate) and write however they want, without any regard to clarity.

Yup - brick's version is better. That doesn't change anything I said.

Nobody has yet answered why a test that aims to judge your ability to comprehend that material you will encounter in graduate school should deliberately choose clear and easy-to-understand passages. It doesn't matter if these scholars are trained in writing - they're publishing work that you need to read and understand in graduate school.

[bigant]

Nobody has yet answered why a test that aims to judge your ability to comprehend that material you will encounter in graduate school should deliberately choose clear and easy-to-understand passages. It doesn't matter if these scholars are trained in writing - they're publishing work that you need to read and understand in graduate school.

Agreed. However, in terms of “realistic testing”, there is not a single text book, paper or anything that I have read (or will read in graduate school) that does not have a title or heading, telling you what the material is about. At the very minimum, and in the interest of source attribution, the passages should have a statement that says, eg "The following passage has been extracted from the paper "The revival of European folk music in the 17^th century."

This still does not excuse the badly written passages. To use a rather loose analogy, you cannot say the user of a Microsoft-based PC is superior to someone who uses an Apple, just because Apple's user interface (UI) is "easier" to use. Agreed that if a place has only Microsoft-based computers, they would want to test you on those computers. If Apple users don’t do well on the test, this doesn't mean Apple users are stupid or aren’t qualified to use computers at all. In the old days, Microsoft had text-based (MS DOS) UI, and Apple had a graphical UI. Apple’s UI would’ve meant far higher productivity, and that’s why Microsoft had to come up with Windows.

[hitchhiker]

Many people here have said that 'scholarly writing' is difficult to understand and hence appropriate for the GRE. But It's obvious that 'scholarly writing' can be totally misleading, look at the example and translation created by the author:

We will experience a strong upturn in our revenues with significant long-term upside potential through a pioneering and aggressive real-estate financial leveraging strategy that dramatically increases home-ownership rates across the nation. The strategy will expand our customer base to include a traditionally underserved and untapped market, subprime customers, who will benefit from the provision of maximum financial leverage with minimal bureaucratic requirements. This strategy will result in the building of tangible asset bases for these customers, and these assets can subsequently function as real estate collateral against further leverage that will stimulate consumer spending, thereby aiding the growth of the national economy indefinitely. Sub-prime defaults will not be a cause for concern because of the introduction of innovative derivative financial instruments. Called Mortgage Backed Securities (MBS), Credit Default Swaps (CDS) and Collaterialized Debt Obligations (CDO), these instruments will serve as risk mitigators and protective hedges against any defaults that occur. In the event of the economy undergoing a downturn, the federal government is expected to perform a guarantor role.

Translation:

New profit-making strategy to rely on giving out bad loans

We will get new customers and earn big money by giving home-loans to people who don’t have jobs, incomes or assets. They won’t even have to put down a deposit or give us any documents.

These people will almost certainly not be able to repay us

.

These risky customers will then use their still-unpaid-for homes as collateral to spend more, getting even deeper into debt. We assume this cycle will continue forever, and will somehow be good for the country.

But since our bank will be at risk, we will sell off these bad loans to investors chasing high returns, or we’ll take out insurance on the loans.

Bottom line: We win big either way – by selling off the loans, or by collecting insurance. If everything goes bust, the government will bail us out

.

Edited September 11, 2012 by hitchhiker

[pemdas]

Many people here have said that 'scholarly writing' is difficult to understand and hence appropriate for the GRE. But It's obvious that 'scholarly writing' can be totally misleading, look at the example and translation created by the author:

We will experience a strong upturn in our revenues with significant long-term upside potential through a pioneering and aggressive real-estate financial leveraging strategy that dramatically increases home-ownership rates across the nation. The strategy will expand our customer base to include a traditionally underserved and untapped market, subprime customers, who will benefit from the provision of maximum financial leverage with minimal bureaucratic requirements. This strategy will result in the building of tangible asset bases for these customers, and these assets can subsequently function as real estate collateral against further leverage that will stimulate consumer spending, thereby aiding the growth of the national economy indefinitely. Sub-prime defaults will not be a cause for concern because of the introduction of innovative derivative financial instruments. Called Mortgage Backed Securities (MBS), Credit Default Swaps (CDS) and Collaterialized Debt Obligations (CDO), these instruments will serve as risk mitigators and protective hedges against any defaults that occur. In the event of the economy undergoing a downturn, the federal government is expected to perform a guarantor role.

Translation:

New profit-making strategy to rely on giving out bad loans

We will get new customers and earn big money by giving home-loans to people who don’t have jobs, incomes or assets. They won’t even have to put down a deposit or give us any documents.

These people will almost certainly not be able to repay us

.

These risky customers will then use their still-unpaid-for homes as collateral to spend more, getting even deeper into debt. We assume this cycle will continue forever, and will somehow be good for the country.

But since our bank will be at risk, we will sell off these bad loans to investors chasing high returns, or we’ll take out insurance on the loans.

Bottom line: We win big either way – by selling off the loans, or by collecting insurance. If everything goes bust, the government will bail us out

.

Let me start from fact that I am not a native speaker of English who spent 19 years on learning the language (since 1993). So I may have no excuse(s) for my general comprehension rather I reserve the right for seldom idiomatic abuse in speech for obvious reasons; I did not speak English as much as the native speakers would for the last 19 years. I am only able to use the language to certain extent - proficient speaker.

Now coming to a point of my post - the simplification you made of the original fragment is not acceptable to me. Yet I agree that it is a very good example of how GRE essay needs to be, i.e. devoid of professional jargon which you removed. However, in the academic realm such simplified writing would not be accepted almost for sure. The translation lacks the sort of abstract thinking - the elements referenced to a few terms and concepts mentioned in the article. The simplified piece looks just abrupt, and the original article allows for some degree of uncertainty of the events discussed.

[hitchhiker]

Now coming to a point of my post - the simplification you made of the original fragment is not acceptable to me. The translation lacks the sort of abstract thinking - the elements referenced to a few terms and concepts mentioned in the article. The simplified piece looks just abrupt, and the original article allows for some degree of uncertainty of the events discussed.

FYI I have not translated the original fragment, it's the author's translation. The issue here is not whether it is acceptable to you or anyone else, the issue here is clarity of communication. The translated version tells it like it is and does not try to obfuscate. The financial services industry has tried to make things look abstract and complex in order to fool people, and people have gotten fooled. I have seen many pieces of academic writing in which it's next to impossible to understand what the author is trying to say, and I sometimes got the feeling that the writers themselves did not know. They simply try to get published one way or another and I know many academics will back me up on this.

[bigant]

Even Stanford knows there's a problem with scholarly writing, that's why they have courses on how to improve writing:

http://www.stanford....ng/lecture1.ppt

http://lane.stanford.edu/classes-consult/laneclass.html?class-id=122

Edited October 4, 2012 by bigant

[spunkrag]

It is funny how this thread has many posts that mimic both the GRE section in question, as well as the article that critiquing the GRE. Grad students never stop obsessing about modeling and simulating the world.

This is somewhat tangential to the (very specific) topic at hand, but in reading over this thread and remembering when I took the GRE (in Nov. of 2002, three weeks after the AW section was added---and programs I applied to didn't bother with the score, preferring to just use writing samples as a much better proxy), the GRE has become a real crapshow. I am aware that the test is used as a screening device for certain very basic skills that are an absolute requirement for success in graduate school; eliminating applicants who fail basic math is essential if AdComs expect to admit students planning on doing, I dunno, maximum likelihood estimation.

I can see a similar view for the reading/writing portions of the exam, but at the same time standardizing language abilities is absurd. And reading comprehension? Please. When was the last time an adviser handed an article to a student, gave instructions to read a few passages, and then abruptly removed the article from the students eyes. Forever. That is not the brand of comprehension required for success as a scholar, and certainly not in graduate seminars, where groups students spend long periods of time analyzing and critiquing research, using as much available data, figures, and context as possible.

I am aware of a couple of top ten programs in my field (it is not the one I list on my profile) that more or less follow university guidelines for GRE minimums, and beyond that consider GRE scores as a predictor of student success to a very minimal degree. Perhaps that is the future of the test?