MGMAT CATs

[tkulkarn]

After taking 3 CATs (3rd under exact test conditions except for the use of the pencil and scrap paper to do calcs, etc.), I am realizing the following:

- in Q I am presented more than 50% 700+ questions and 40% 600+ ones. I ran out of time and had to guess on the last 6 (all wrong). Nonetheless 45+ final score
- in V, the situation does not change much either. I am presented 40% 700+ questions and I get 40%+ of these right. Ran out of time and had to guess the last RC passage also.

I have not faced this problem with GMAT Prep where I do not have a breakdown of the the difficulty levels of the questions. But I did not have timing issue with it.

Would the GMAT actual be closer to GMAT prep or MGMAT? If the amount of reading I need to do is similar to the latter, then I am going to have time management problems in V.

Thank you.

[StaceyKoprince]

It can vary a little bit depending upon the particular mix of questions you hit. As far as we can tell, the GMATPrep database of questions is a bit thinner than the real test (or ours), which means you may get fewer really mind-numbing questions if you're at the higher end of the scale (which, of course, translates into a time advantage). (Not definitely, just maybe.)

We tend to go for the more pessimistic view on our tests - overpreparation is better than underpreparation (in our opinion!). In general, most people find that they have to give up on some questions at some point throughout a section, no matter how good they get. The question becomes: do you want to be forced to give up on a bunch in a row at the end, or do you want to be able to choose the ones on which to give up as you move through the section? Hopefully, the answer to that is obvious. :)

So, do assume that you are sometimes going to hit things on which you are going to be tempted to spend too much time. Part of your task is to identify when spending that extra time would be a waste of time, and - literally - get the question wrong faster. (That is, if you're going to get it wrong anyway, get it wrong in 2m rather than 3.5m. **) If you don't make these calls along the way, then you're more likely to end up in the situation you've described: having to make random guesses on several in a row at the end. (And that's almost the worst possible scenario.)

** I'd also, by the way, prefer to get something wrong in 2m rather than get it right in 4m. That sounds bizarre, doesn't it? Wouldn't I always prefer to get something right?

No, actually. Let's say I'm doing a quant question. 4m is 2m too long for quant. Now I'm 2m behind. Most people will then try to speed up on questions that they know how to do in order to get back on track. How will you try to save that 2m? By spending, say, 30s less on 4 other problems that you know how to do. "Problems that you know how to do" = lower-level problems. Rushing = more likely to make a careless mistake.

So I'm going to give myself 4 chances to make a careless mistake on a lower-level problem in order to get just one other problem right? I don't think so. (Especially when I remember that if I do get that impossible 4m question right, what's my reward? An even harder question! No thanks.) :)

[tkulkarn]

Thank you Stacey.

I am done with my GMAT for now. I took it end of December'09. got into the 90th percentile overall. I read at many other forums that MGMAT CATs over rather than under-prepare you. Especially the first few are way trickier. GMATprep and GMAT's algorithms are truly adaptive. When I took practice test using GMATprep, I found that the algorithm would adapt and not throw one question at me harder than I could solve.
My last MGMAT CAT score was closer to my GMAT score but my first mgmat cat score depressed me quite a bit.

It can vary a little bit depending upon the particular mix of questions you hit. As far as we can tell, the GMATPrep database of questions is a bit thinner than the real test (or ours), which means you may get fewer really mind-numbing questions if you're at the higher end of the scale (which, of course, translates into a time advantage). (Not definitely, just maybe.)

We tend to go for the more pessimistic view on our tests - overpreparation is better than underpreparation (in our opinion!). In general, most people find that they have to give up on some questions at some point throughout a section, no matter how good they get. The question becomes: do you want to be forced to give up on a bunch in a row at the end, or do you want to be able to choose the ones on which to give up as you move through the section? Hopefully, the answer to that is obvious. :)

So, do assume that you are sometimes going to hit things on which you are going to be tempted to spend too much time. Part of your task is to identify when spending that extra time would be a waste of time, and - literally - get the question wrong faster. (That is, if you're going to get it wrong anyway, get it wrong in 2m rather than 3.5m. **) If you don't make these calls along the way, then you're more likely to end up in the situation you've described: having to make random guesses on several in a row at the end. (And that's almost the worst possible scenario.)

** I'd also, by the way, prefer to get something wrong in 2m rather than get it right in 4m. That sounds bizarre, doesn't it? Wouldn't I always prefer to get something right?

No, actually. Let's say I'm doing a quant question. 4m is 2m too long for quant. Now I'm 2m behind. Most people will then try to speed up on questions that they know how to do in order to get back on track. How will you try to save that 2m? By spending, say, 30s less on 4 other problems that you know how to do. "Problems that you know how to do" = lower-level problems. Rushing = more likely to make a careless mistake.

So I'm going to give myself 4 chances to make a careless mistake on a lower-level problem in order to get just one other problem right? I don't think so. (Especially when I remember that if I do get that impossible 4m question right, what's my reward? An even harder question! No thanks.) :)

[StaceyKoprince]

Congrats - great job!

Especially the first few are way trickier.

There's no difference between the earlier and later MGMAT tests that you take. The exact same algorithm is used for every test. The full question database is the same for all tests, as well - we don't give especially tricky questions early on or anything like that.

You just got better - that's why it seemed less tricky as time went on. :)

Good luck with your apps!