ThomBoh wrote:I'm inclined to disagree. To say that there was a question on a particular topic, without saying what the actual question was (e.g. saying there was a question about quantum efficiency of a detector without saying exactly what they were asking you to calculate and in what scenario) shouldn't really be a violation. By that logic there's a lot of information about what types of questions have appeared in the past in the Conquering the Physics GRE book that would be considered a violation and if ETS cared they would take action.
I agree with you that I also personally do not think that saying "There was a question about the quantum efficiency of a detector" is necessarily a violation of the ETS agreement. I think this should be okay because this is not enough information for someone else to determine what question was actually being asked. However, if the discussion had gone further and people discussed exactly what was asked, how to solve it, what the answer was etc. then this would certainly be a violation.
What I was trying to say though, is that the administrator ended up editing out all references to any questions, without regard for how much or how little they revealed. I think this was necessary because it's easier to remove all references rather than determining exactly which ones would be a violation and in cases where the line is blurry, it's probably better to err on the side of caution and not discuss it.
I don't normally agree with "censorship" but in this case, I think it is justified because the gains one might have from discussing the questions is very small. But that's only my opinion of course.
Pointing out the types of obscure topics that showed up on my exam (without discussing what the question on the topic actually was) is really one of the more useful things this forum could provide: there was no way I would have thought to study those types of things because they specifically did not show up (or showed up a negligible amount) on the published practice exams and weren't mentioned in any course I've ever taken, in my prep book, or in ETS's topics list. It's important for people to know that their exam could either be very similar in content and style to practice exams as was the case with my september exam or could be very different as was the case with my October exam.
In my opinion, I don't think that's the most useful aspect of this website. I think the most useful portions are the parts that relate to all aspects of a graduate school application, not just the PGRE. For example, the annual profile and results postings.
In addition, obscure/random questions are by definition, unpredictable and knowing what obscure questions appeared on one edition of the test doesn't give you very much predictive power to guess future editions of the test. That is, I don't think knowing that your test had a question about quantum efficiency of a detector (I didn't write this exam so I have no idea what the question actually asked) would help me prepare for a PGRE if I were to take it in the future.
I actually believe that ETS does give us more than enough information to prepare for the test. We have the approximate breakdown of topics and I think they are fairly accurate, within reasonable "error bars". That is, ETS says 20% of the questions are Classical Mechanics. I don't think I can get enough information from this forum to be confident that the real tests are actually 35% Classical Mechanics (for example) because 1) no one has posted the breakdown of their individual test and 2) not enough people post enough information to distinguish "signal" over the random "noise" of test-to-test variations. In addition, even if the real percentage is like 15% or 25% Classical Mechanics, it's not like this information would significantly change study strategies.
You also mentioned that some of the questions didn't appear on ETS' topic list. They do say that "nearly" all questions will be on their list. Depending on what was actually asked in the "quantum efficiency" question, it could easily fall under "Laboratory Methods" or "Specialized Topics" (my exposure to this concept in my studies have all been related to how astronomical CCDs work). These two sections are highly unpredictable, and they make up a good chunk of the test (15%). Thus, this gives ETS a lot of wiggle room to put things test-takers might not expect in their tests, so I don't think it's very useful to know breakdowns of subjects down to the exact # of questions. Honestly, I think the only significant difference in the percentages given is that Classical Mechanics and E&M are the two categories with about twice as much weight each as any other category.
Finally, you say it's important for people to know whether or not the questions can be very similar to the old exams or very different. But that's basically all the possibilities -- all exams will either be like the old ones, or not like the old ones. It's not useful to know this unless you can somehow provide enough information that people will know what the April 2014 exam will be like. It's important for you to point out that exams are not always like the practice ones but I feel like this is already known. After all, ETS is pretty careful to say that you cannot expect the real test to be like the practice exams published. But if I am wrong about what is generally known, it's still possible for people to convey the important message that the latest test is extremely different from all previous tests without discussing the questions involved!