MGMAT exams

raulverde · Posted: Tue Aug 28, 2007 6:21 am Post subject: MGMAT exams

Guys,

I am really confused abt my MGMAT results. My scores are dropping -
Exam 1 - 680
Exam 2 - 670
Exam 3 - 650

All exams were timed. The problem is that i believe i was getting better at GMAT in between these exams. My success rate on OG problems is abt 90%.

I am absolutely positive that i can easily get 700+ on MGMAT if i don't time the exam...but that would defeat the purpose.

The i see posters on forums who get 750 - 760 on MGMAT exams...and i say to myself...WHAT THE HELL ?

I am scared to even take the GMAT PREP exam coz i feel i am not prepared for it untill i get a 690 on MGMAT.

Anyways....that was my rant

paradox · Posted: Tue Aug 28, 2007 6:12 pm Post subject:

Hmm, yeah, I'm not sure what to make of my scores either, and my scores are going down, but I am enforcing stricter timing as well. I got a 530 on my last MGMAT CAT...had a bit of a brain freeze. My timing seems way off. I set the timing to 3min/question and got about a 670 on one of the MGMAT CAT exams (2nd one I believe).

This may be specific to me only, but I don't buy the argument that if you can't do it in 2 mins, you can't do it at all...extra time seems to bump my score way up.

As another reference point, I got a 690 on the 1st GMAT Prep exam (doing it a second time) a while ago. So I am totally confused as to what I can expect on the real GMAT which is in about a week. I plan on taking the 2nd GMAT Prep exam in a few days and will use that as my best estimate.

paradox · Posted: Tue Aug 28, 2007 6:17 pm Post subject:

BTW, I too am in awe of these posters who get 750+ on MGMAT exams under real timed conditions. I wonder if the ambient temperature of the room rises by any noticeable measure when they take the tests. Wish that was me (or is it were me? Very Happy

)

beny · Posted: Tue Aug 28, 2007 6:29 pm Post subject:

Here's my take on MGMAT:

The company was founded by an intelligent, high-achieving individual, and recruits intelligent, high-achieving individuals. Therefore, the company is biased in that it assumes that all of its clients are also intelligent, high-achieving individuals. That's why some people rave about the program, while others are less enthusiastic. I think MGMAT is probably a good program for people scoring 690+, but not as useful for people scoring less.

Regarding the timing... just wait until you take the real GMAT. One thing that is great about MGMAT was that they teach you not to become married to a problem. If you can not solve it in 2 minutes, it's much more worth it to make an educated guess. If you spend 3-5 minutes trying to solve the question on the real test, it will hurt you much more than it will help you. You'll still get a majority of these questions wrong, and you will run out of time on your ending questions.

The problem with MGMAT is that not all the questions are indicitive of actual GMAT (especially for lower-level questions). IMO, while some of the 700+ questions (including quant) may actually represent what's on the test, some of the questions that are labled 500-700 are more difficult than what's on the test at those levels. I think this question bank bias is because of the people running the company and the clients that they expect.

Another thing to consider: Retaking a GMATPrep test will bump your score up a lot... even if you are only seeing 3-5 repeating questions per section. These questions, in addition to being free points, take no time to answer, thus give you more time to answer the rest of your questions... and an additional 6-10 minutes is quite a lot.

One of the biggest problems I had on the test was with timing. While my math skills are usually strong enough to answer any question (given sufficient time) I still had to guess on at least 20% of the quant questions on the test. While I knew that I could have solved these questions correctly if I had infinite time, I also knew that, at a certain point, it would have been much smarter to make a guess.

The very first question I got on the GMAT was a combinations problem. I looked at it, thought about it for 20 seconds, then took a random guess and moved on. Most people obsess over the first x problems, since these are the "most important", but I knew, timing wise, it didn't make sense to invest that much time in it. Even though I got this question wrong, I was able to score well on the quant, and ended up with a high score.

I found the GMAT quant harder than GMATPrep quant, and, if I had relied solely on GMATPrep, I would have been too attached to the problems to guess and move on (I was able to answer all GMATPrep questions easily, with about 5-7 minutes to spare at the end of the section and would have expected this on the actual GMAT as well). I am grateful that I took a MGMAT CAT before the test, because it did teach me the importance of timing.

beny · Posted: Tue Aug 28, 2007 6:32 pm Post subject:

I didn't think getting 750+ on MGMAT was that difficult... I got a LOT of questions wrong, and I didn't know how to answer a LOT of questions (guessing on maybe 1/4 to 1/3 of the quant questions), and I still managed a bunch of 750+'s under timed conditions. I think the MGMAT grading algorithm is extremely generous...

beatthegmat · Posted: Tue Aug 28, 2007 8:27 pm Post subject:

Try not to read too much into your practice scores--they are still strong, and they may not be an accurate indicator of your actual performance. It's always best to focus more on analyzing your errors/weaknesses and systematically addressing them.

I would think about taking the second GMATPrep--it's the most accurate indicator of your actual GMAT performance and will be the most telling evidence of your progress.

Good luck and don't stress too much!

beatthegmat · Posted: Tue Aug 28, 2007 8:28 pm Post subject:

One more note--it sounds like need to work on your timing. Check out http://www.beatthegmat.com/wiki/ to explore some strategies on how to work on timing and getting comfortable with GMAT test taking conditions.

raulverde · Posted: Wed Aug 29, 2007 6:35 am Post subject:

thanx for the suggestions guys. I believe my problem is spending too much time on a particular problem if i "feel" i can solve it. That definitely screws me up in the long run.

Will checkout the wiki for timing issues.

gracias

paradox · Posted: Wed Aug 29, 2007 12:36 pm Post subject:

Beny, thanks for your post - good info there.

I took the final MGMAT test and scored a 680 under timed conditions. I've been out of school for some time and have never taken a CAT test before, so my brain is a bit rusty on the material and test taking in general.

I tried to take your advice and to be able to let go of a question, and it seems to have helped...in fact I finished the verbal section early! Also, I tried to ignore the "water torture" of the clock ticking away - I realized that part of my concentration was consumed with that, instead of the question at hand. I still need to get better at these two things though. Lastly, I provided a somewhat legible answer in AWA instead of just skipping it, and I think it actually helped warm me up and focus my concentration a bit.

Here is a listing of my MGMAT scores for those interested:

1 - 530 (timed)
2 - 700 (/w extra time)
3 - 670 (/w extra time)
4 - 590 (timed)
5 - 520 (timed)
6 - 680 (timed)

Taking a 1 day break between 5 & 6 may have helped too (I did 4 and 5 on conscutive days).

My plan is to focus on improving my Quant score in the short time I have left.

DrivinWest · Posted: Sun Sep 02, 2007 11:50 pm Post subject:

I'm another MGMAT 9-week class taker and I'm 8 days from my GMAT. Between now and then I've scheduled two GMATPrep practice tests and lots of review. Over the last few months I've taken 6 TIMED MGMAT CATs with the results of the last 5 below:

620 - Q42 (66%), V33 (69%)
650 - Q43 (70%), V36 (81%)
710 - Q44 (74%), V42 (96%)
700 - Q46 (79%), V40 (91%)
700 - Q46 (79%), V40 (91%)

with the latest 700 being today's score. I'm mostly content with my Quant score - MGMAT Quant is supposed to be harder than the real thing but I'm looking to attack a few weaknesses there. That said, I've been studying my butt off but haven't seen any quantifiable improvement on the last two exams - my confidence has improved, however.

Verbal also hasn't improved which is most frustrating. Running analysis reports on my last two test confirms that CR is my weak Verbal spot. CR!? It seems that some people think the MGMAT CR passages are a little iffy. I sure hope so. I'm getting less than half of them right on MGMAT CATs which give me an abundance of 600-700 and 700-800 level questions. If I can just get my Verbal in the 42-44 range I'll be where I want to be.

The OG CR sample problems do seem easier...

-Aaron

Stacey Koprince · Posted: Mon Sep 10, 2007 12:22 pm Post subject:

Great discussion, guys. FYI: standard deviation on our tests is about 50 points. SD on the real thing is about 30 points. GMATPrep's SD isn't published, but we know it's going to be higher than the real thing, and I just assume it's also going to be better than anything a prep company can do - so I think of it as 40 points. (Though that is totally just a guess on my part.)

I have had a lot of my higher-scoring students tell me that they think GMATPrep is easier than the real thing. I think this is because the question pool for GMATPrep is pretty shallow (that is, there aren't that many questions), so if you score very well, you essentially use up your high-level question pool and the test then feels easier than the real thing b/c the real thing will have TONS of really hard stuff to keep throwing at you.

Also, re: the "it doesn't help you to take a lot more time" issue - academic, scientifically valid studies have been done to show that, on average, the more time you spend about about 2min40sec, the less likely you are to get something right. Plus, that extra time has to come from somewhere, and typically it causes people to make careless mistakes on lower-level problems they should have gotten right but didn't because they were rushing.

Remember that, on the real thing, the questions are all written such that someone can do them in 2 min. Maybe not you, but someone. If you feel like you need 4 min to do such a problem, that's actually an indication that you don't really know how to do it b/c you can't do it in the expected 2 min timeframe.

Particularly on the math, you have to remember that, no matter how good you are, the test will give you stuff you cannot do in 2 minutes, and it will also give you stuff you cannot do at all (without, say, a textbook or an instructor sitting next to you Smile

). If you get sucked into spending too much time on these kinds of problems, you will have a double whammy - you'll be more likely to get that problem wrong, plus you'll likely cost yourself points elsewhere by having to rush.

I agree that some of our quant questions are too computation-intensive. We actually have an analysis algorithm we run on our own test questions as we get enough data points from students, and we are going back in and altering or dropping questions that take too long, are too easy or too hard, that aren't properly discriminating (that is, people with lower scores should be more likely to get a particular question wrong and people with higher scores should be more likely to get it right), etc.