Physics GRE - optimum # of questions to attempt?

[dipole]

Just to put this question in context, I have been preparing for the GRE for months now, I've been developing a strategy to maximize my score regardless of what my knowledge of physics is or isn't. I'm not asking for basic advice.

The only part of my analysis that is lacking is trying to determine the optimum number of questions to attempt, or put another way, how many questions should you outright skip completely with no attempt at trying.

I do not believe the optimum number is all 100, for several reasons. First, you should focus on your strengths, and topics you're already weak on to begin with are not worth wasting time with (since you're more likely to get it wrong anyways). Second, you only need a raw score of about 65 to get a good score on the exam - that leaves room for skipping quite a few questions (though you do need some cushion for getting questions wrong). Skipping some questions means you have more time to work on those you feel you know how to do, and means you're less likely to make a mistake.

I would guess that if you plotted your average raw score (assuming you had many exams to take and took them far enough apart that taking one didn't influence your score on another), versus the number of questions you actually attempted (without randomly guessing), then it would peak somewhere around 80 questions attempted. That number just comes from my own feeling of doing the exams and knowing what doing 80 feels like, and it seems like a good number. I could be wrong though, it could actually be lower, or maybe higher. Anyone have thoughts or opinions on this?

 

[TMFKAN64]

I don't think it's useful to think in terms of how many questions you should skip.

Given the time constraints, you have a little less than two minutes per question. Priority #1 should be to at least *see* all of the questions, so I think you should work through the test, spending at most two minutes on any question you think you can solve in that time and skipping any question you think will take longer than that. If you come up with an answer that doesn't match one of the choices, skip it and move on.

Once you have caught the low-hanging fruit, go back through the test a second time to work on any problems you worked on but didn't complete during the first pass.

If they haven't called time, try the problems you totally skipped.

I think the optimal algorithm is more a question of triage than selecting a number of questions to skip initially.

 

[dipole]

I don't think it's useful to think in terms of how many questions you should skip.

Given the time constraints, you have a little less than two minutes per question. Priority #1 should be to at least *see* all of the questions, so I think you should work through the test, spending at most two minutes on any question you think you can solve in that time and skipping any question you think will take longer than that. If you come up with an answer that doesn't match one of the choices, skip it and move on.

Once you have caught the low-hanging fruit, go back through the test a second time to work on any problems you worked on but didn't complete during the first pass.

If they haven't called time, try the problems you totally skipped.

I think the optimal algorithm is more a question of triage than selecting a number of questions to skip initially.

Yes I totally agree and this is a big part of my strategy, but even questions you know exactly how to do can sometimes take you longer than 1.7 minutes for whatever reason (maybe you dropped a factor of two somewhere or whatever). To me it seems that you should really give yourself a little more time than 1.7 minutes for questions you know how to do, and less time for those you don't. This means skipping some you don't know or are unsure about.

For example, you see a question on RLC circuits and you can't remember how inductors work. All the answers are numerical. You have almost no way of eliminating any choices - you should just skip this question and not even bother coming back to it. That extra minute is better spent on block and wedge problems that are easy but might take you slightly longer than the average question.

 

[TMFKAN64]

See, I don't think you should spend 3 minutes on *any* question until you've at least seen them all. Otherwise, all of the "just *one* more minute" questions will add up, and you could easily run out of time before you get to the end of the test.

This requires a lot of discipline... it's tough to give up (even temporarily) on a question that you *almost* have!