Stagnated CAT Scores---Please Help!

[jesse.a.rosales]

Hello,

My test date is May 13, 2014 and I have just over 3 weeks until my exam. I am aiming for a Q44V37 score- a score which I feel is achievable given the questions I am getting during the practice tests. I would particularly like some advice from those who have successfully addressed the following issues: timing, skip questions and concept mastery. However, any help from those who have overcome similar score stagnation would be welcomed.

I began my studies with an MGMAT in-person course that ran Jan-March. My CAT practice exams have had the following breakdowns:

1/26/14 CAT1 (w/o IR or essay): Q37V28 540
3/1/14 CAT2: Q37V24 520
3/2/14 CAT3: Q36V30 550 <--- this is when I first ID'd my timing issue and tried to address it.
3/22/14 CAT4: Q29V31 510
4/6/14 CAT 5: Q36V29 550
4/19/14 CAT 6: Q37V31 570

Study Background: I've had an error log since February but recently started outlining source of error, applicable rules for incorrect answers and tips on how to avoid it in the future. Admittedly, I haven't spent too much time reviewing the error log once I have already thoroughly evaluated incorrect ?'s. I have begun to ID skip questions. I've created flash cards for all notes but not yet for troublesome incorrect questions.

General characteristics of CAT's 1-5: my timing was all over the place, I had a couple 4min+ questions, many 3 min, and I was usually off benchmark times from 3-8min. On both sections I had streaks of 4+ incorrect questions regularly.

Since CAT 5 I have focused on timing and performing all practice problems within general time benchmarks (2min for quant, 1:15 for SC, 2:15ish for CR and ~7-8min for RC sections). As a result, on my CAT6 the timing has become more normalized with only a few 3min+ questions in either of the two sections and shorter streaks of incorrect questions as before.

The timing strategy I have adopted for Verbal is to spend a little bit more time on the first 12 questions to increase difficulty level and attempt to maintain from then on. Quant is being approached as all questions being equally important.

My study strategy since CAT 5 has shifted from learning the content to performing 12-15 OG practice problems about 4-5/week. My time breakdown for quant:verbal has been about 80:20, respectively. I'm pretty diligent about getting at least 2hrs in/day 6days/week.

CAT6 Quant Stats:
#15 I was +2min/60th%
#23 I was +1.5min/54%
#30 I was +0/44%
#37 42%

CAT6 Verbal Stats:
#12 I was +6min/84%
#27 I was +10min/69%
#32 I was +8min/75%
#38 I was +2min/66%
#41 62%

What are some study techniques you think I should be embracing for the test and study time?
Should I focus on bringing up my verbal score to compensate for a flailing quant score?
Any help is greatly appreciated.

Thanks,
Jesse

[StaceyKoprince]

Your most recent test has Q37, or 42nd percentile; your goal is Q44, or 61st percentile, a jump of nearly 20 percentile points. That's a decently significant jump - you may need more than 3 weeks to achieve that.

Your most recent verbal score is V31, or 60th percentile. Your goal is 37, or 82nd percentile, again a jump of about 20 percentile points.

Most people would need more than 3 weeks to achieve that kind of jump in both sections.

Do you have a particular deadline? Or is that the test date you chose for yourself a while back? If you don't have a particular deadline, you may want to consider postponing the test. You could also take it with the assumption that this is a dry run and you'll take it again. (Then, if you don't need to, it's a bonus / nice surprise.)

If you have a deadline, you'll just have to do what you can in the time that you have.

The timing strategy I have adopted for Verbal is to spend a little bit more time on the first 12 questions to increase difficulty level and attempt to maintain from then on.

If you need more time than average to get more of those earlier ones right, then how are you going to maintain that level with less average time going forward (for the rest of the section)? This is not a good idea. (And notice that it didn't work on your test! You were at 80th+ percentile early on, but at 62nd by the time you finished.)

All of the questions are inherently "worth" the same amount. The only ones that don't count quite as much are what are called your "outliers" - the ones that are at the far end (high OR low) of your scoring scale. So spending extra time to get an outlier (hard-for-you Q) right is actually a waste of time because that question won't count as much toward your score!

Plus, you will not be able to maintain that higher level with less average time available. Do what you can do, and let the ones that you can't do in a reasonable amount of time go.

There isn't a way to "game" the algorithm. Believe me - I've been looking for 20 years. :)

It's good that you recognized your timing problem and have been working to make it better. I'd like you to dive into the data a little more deeply and make sure you've actually been making wise decisions as you fix your timing.

When people first realize they're messing up the timing, they will often default to a certain pattern that isn't actually much better than the old one. I'll use quant as an example. People want to hit the 2-min average so badly that they cut themselves off on problems that are actually worth spending 2m15s or 2m30s on.

Next, they'll choose to spend extra time on problems that are still giving them that "if I could just have a little more time, I know I can figure this out!" feeling. Those are the ones you want to cut off *faster.*

Instead, give yourself permission to spend an extra 15 or 30 seconds on a problem that you know you know how to do - you have the plan, you can see your way to the end, you've got this. The calculations are just a little annoying or it's a longer / harder problem, so it takes a bit longer than average.

Also, people will often speed up a bit on problems that they do know how to do and are maybe even already answering faster than the average. Do NOT speed up on these - all you're doing is giving yourself a chance to make a mistake.

Those ones that you were spending 3 or 4+ minutes on - those are the ones to cut off by 1 to 1.5m.

If, on the other hand, I finish a problem and I think I haven't even spent a minute on it yet, I check my work. Just to make sure. :)

Also, I only bail on a problem within 15 seconds if I know that it is one of my absolute biggest weaknesses. Otherwise, every problem gets a read-through / at least about 30-45 seconds. At a minimum, I may be able to figure out how to narrow down the answers.

If I'm still totally clueless by 30-45 seconds, though, then I guess and move on.

So, go through your last test problem by problem. Did you actually make good decisions about where to spend your time? Or can you improve that aspect of things?

There's one more important thing for you to know - so important that I saved it for last because I want you to think about it. The GMAT is what is called a "where you end is what you get" test. Stop trying to lift your score high earlier on at the expense of your performance later on, because where you end the section is what you get.

Save yourself that time - and, equally important, that mental energy - to spread evenly throughout the section / test. If you bang your head against the wall too many times on hard questions, you're going to miss easier ones later in the section or test because your brain can only do so much. Every choice you make it not made in isolation; it has an impact on your performance for the rest of the test.

I will actually sometimes see a problem that I know I can do, though it'll probably take me full time if not a bit longer, but I don't do it anyway because I know it will take more brain energy than I'm willing to spend on a single problem!

If you would also like advice on specific areas to study, use the below to analyze your most recent MGMAT CAT (this should take you a minimum of 1 hour):
https://www.manhattangmat.com/blog/inde ... ts-part-1/

And if you haven't already read the below two articles, do so:
https://www.manhattangmat.com/blog/inde ... lly-tests/
http://www.manhattangmat.com/blog/index ... -the-gmat/

Figure out your specific strengths and weaknesses as well as what you think you should do based on that analysis. Then come back here and tell us; we'll tell you whether we agree and advise you further. (Note: do share an analysis with us, not just the raw data. Part of getting better is developing your ability to analyze your results - figure out what they mean and what you think you should do about them!)

[jesse.a.rosales]

I appreciate you taking time to respond, Stacey.

I set the test date with a mountain climbing trip and application deadline in mind. I can push the test date as far back as May 27th but anything beyond that would require me to cancel my trip which I may do if necessary.

My current plan is to see how close I can get to my score goal between now and May 3 and then make a decision on whether or not to push back my test date. May 3 is when I take my GMATprep CAT1 and it would be the day I would switch to refining/practicing my strengths and test day game plan... but I'm open to suggestion regarding this approach.

Here are the stats I've put together on my CAT6 Verbal.
-------------Total Time--AvgTime-#correct--Total#
CR Avg #1-20------21:19----2:39-----5-----8
CR Avg #21-41----10:18----1:43 -----3-----6
SC Avg #1-20------8:11-----1:38 -----1-----5
SC Avg #21-41-----8:27-----0:56-----5-----9
RC Avg #1-20------17:44----2:57-----3-----6
RC Avg #21-41-----8:25-----1:24-----3-----6

I broke up the results of my CAT6 by 1st half and 2nd half of Verbal and your point about each question being equally important is made clear through the data above. I was spending, in some cases, twice as much time on questions in the 1st half compared to those in the 2nd half. This hurt me because I should've gotten them wrong anyway AND they took away time from the 600-700 level questions I would've had a better chance of getting correct later on. Even though I transitioned to a normal per question pacing about half way through, I was still able to get back to the 75th percentile 3/4 way through the Verbal section; I attribute this to correctly answering 600-700 level questions. However, at the end I was forced to get back on time by rushing 600-700 level questions and I even had to guess (incorrectly) on 3 easier RC questions (questions that I would normally answer correctly).

I've had poor accuracy on 700-800 level questions on CR, RC and SC over the past 3 CAT's in which I've had 33%, 50%, 25% accuracy, respectively. In contrast, 600-700 level questions in CR, RC and SC have been 47%, 67% and 71% accuracy, respectively. By spending too much time on early 700-800 level questions, i'm effectively shooting myself in the 600-700 level questions' feet. My 600-700 level weakness is CR and that is probably my verbal area weakness in general. Similarly, the accuracy of my RC subcategories are above 70% except for inference, which is at 33%- and that is closely related to CR.

Weaknesses over last 3 CAT's (*true for CAT6):
SC: Parallelism 33%* and 1:45 time, Pronouns 50% and 1:47
CR: Weaken the A 33%*, explain the D 33%, Find the A 43%*, Describe the R 40%
RC: Inference 33%*

Strengths over last 3 CAT's:
Maybe RC? 58% but the timing is still a little long

I know I need to improve CR in the 600-700 level immediately but what is the most effective way to improve the accuracy of 700-800 level questions across all categories? I know how to deconstruct arguments relatively well (it just takes me ~60-75secs to do so and I end up being rushed), I can cite most of the SC rules (at the very least, they make sense to me when I see the correct answer). I have not been keeping an error log for any of the verbal questions, instead I've been relying on understanding CR procedure and SC rules- no work has been done on RC. What habits should I be changing to get better results? I know I need improve my ability to ID what the SC question is testing and cite all SC rules but at this point am I better off doing more practice problems and working on evenly dispersed timing? Mind you, up to this point I have spent about 80% of my study time on Quant...


CAT6 and OG Quant Analysis:

The ratio my incorrect responses for Careless Errors: Concept Mastery: Concept Errors is 1:1:2. "Careless errors" are arithmetic miscalculations. Concept Mastery means I know how to do it but didn't have enough time or the vision. Concept error is I seriously went awry. Additionally, I usually catch careless errors and concept mastery in my 2nd pass without looking at the answers.

I've read "The Second Level of Learning" article and all of that great info seems useful to address some of the general themes above. As I mentioned in my previous post, I do not spent a tremendous amount of time reviewing my error log after I analyze why I got a question wrong the first time. At this point, would I most benefit from redoing all my incorrect ?'s and applying that 2nd Level Learning procedure or should I only apply that 2nd level learning from here on out? Any other tips on avoiding careless mistakes? I used to have problems understanding what the question is asking but I have made progress in this area by slowing down and reading the question only once.


Weaknesses over the last 3 CAT's:
Number Properties 25%(all areas except for statistics), Combinatorics 25% and 2:33, consecutive integers 33%, percents 11%, prime factorization and divisibility, overlapping sets (25%), Rates and work (43%), linear equations (43%)


Analysis of last 3 CAT's:
Problem Solving 66 29 37 0 44% 1:46 2:06 550 640
300 - 500 7 6 1 86% 1:07 3:45
500 - 600 28 15 13 54% 1:46 1:36
600 - 700 20 7 13 35% 2:11 2:11
700 - 800 11 1 10 9% 2:51 2:30

Data Sufficiency 45 20 25 0 44% 1:59 1:57 570 640
300 - 500 3 3 0 100% 1:11 NA
500 - 600 15 8 7 53% 1:50 1:51
600 - 700 23 9 14 39% 2:24 1:59
700 - 800 4 0 4 0% NA 2:01

I have a tentative test date of May 13. What is my best approach to improving my quant? What else would you recommend for me to achieve my goal? Thanks!

[StaceyKoprince]

That sounds fine (deciding on May 3rd).

So your analysis shows that you do have some timing problems that are bringing your score down. That is, your ability level is already higher than your scoring level. Fix those timing problems and your score will lift to where it should be.

what is the most effective way to improve the accuracy of 700-800 level questions across all categories?

Stop trying to. Your goal score is what - high 600s? 700? You don't need to get *harder* questions right. You need to make sure that you're not missing many sub-700 level questions.

On verbal, do sets of mixed questions. Analyze both the questions and your thought process. If you get an SC right for the right reasons, then don't worry about being able to officially name that rule. If you get it wrong, or get it right but it was really a guess / you weren't sure why (or you got it right for the wrong reasons), then go into your book and learn that rule (or those rules).

For CR and RC, see whether the problem lies with the argument / passage (the text) or with the question / answer choices. Either way, figure out why you made whatever mistakes you made or missed understanding whatever you needed to understand. That will tell you what you need to do next time to get better.

The ratio my incorrect responses for Careless Errors: Concept Mastery: Concept Errors is 1:1:2.

In the short time you have left, ignore #3 and concentrate on #1 and #2. How can you minimize careless errors?
http://www.manhattangmat.com/blog/index ... -the-gmat/

What stymied you on the Concept Mastery? What do you need to learn so that you know how to do that next time?

Do mixed sets of questions, incorporating some old ones (including ones where you made mistakes in past) but also incorporating new ones. Then ANALYZE those problems to death. Don't do hundreds of problems - the analysis is by far the most important step and should take time.

Number Properties 25%(all areas except for statistics)

Did you mean to say Word Problems? Statistics shows up in Word Problems, not in Number Properties.

Dumb combinatorics - get them wrong fast. You'll probably see just 1 on the test. Ditto consecutive integers at this point.

Do dive into Percents and linear equations - those are very common.

Get those 700+ PS questions wrong faster and use that time elsewhere. You had 15 700+ quant questions on the last 3 tests and you answered just 1 correctly. If you guess blindly on 15 questions, you'll have a 1 in 5 chance of answering correctly - or about 3 questions!

I don't know your goal score, but my guess is that you are going to need more than another week to hit your goal. Work on the things I mentioned above - be VERY opportunistic. Do NOT try to do / learn everything. You have limited time, so you have to prioritize - which means: work on timing, careless errors, and other "low-hanging fruit" that will give you the biggest bang for your buck.

[jesse.a.rosales]

I took the test today and received a 490 Q32V25 (5.13.14). Absolutely shocked considering my most recent GMATprep practice test was a 610 Q39V34 (5.9.14).

Overall:
I was well rested going into the exam. I had some nerves on the cab ride over and while sitting in the waiting room (they had me wait for 30 min as people were called in one by one) but I felt as if the nerves were calmed by the time I made it through the AWA and IR sections- I ended up with a 5 in IR and I felt as if I had put together a strong AWA. Admittedly, I put a little bit more focus on these sections on test day than I had on any practice exam but other than that, everything else was fairly similar- I got a good night's rest, hydrated, etc.

Quant:
I felt as if I had started off strongly being that I was seeing some questions that were challenging for what I have established to be my skill level (moderate OG13 questions). I was focusing on ditching excessively tough, time consuming problems and implementing answer choice elimination. I felt as if I had performed equivalent to how I did on my most recent GMATprep practice test. I felt confident about my performance going into the break. I stayed on timing benchmarks throughout this section. I recognized about 3 seemingly identicle problems from my practice and I utilized answer choice elimination (spotting parallel info) on early DS problems I found to be time consuming/challenging.

Verbal:
I felt uncertain about my first 3 questions (I wasn't confident about the 2 SC splits or CR answer) but I was confident for the remainder of the section. I was seeing some very challenging SC questions between #20 and 30 and I was relatively close to my timing benchmarks until ~25. I spent too much time on an SC and CR question in the late 20's and this threw me off time a little and forced me to ditch two time consuming CR questions at 32 and 39. I felt relatively confident otherwise.

Final 3 weeks leading up to the exam:

My final two GMATprep practice tests were as follow
5.3.14 GMAT Prep1 540 Q38 V26
5.9.14 GMAT Prep2 610 Q39 V34

After these practice tests I focused on analyzing timing issues, implementing answer choice elimination, identifying careless errors, and identifying strengths. I then took that analysis and applied it to sets of ~20 questions/day (alternating between Verbal and Quant every other day) that were focusing on my strengths and most frequently tested topics. I made ~200 flashcards over the final 3 weeks and superficially reviewed those for 2hrs on the final day before my exam. My flashcards were "when I see____ I need to do____ to avoid" type cards. By the end of my preparation I was able to ID trap answers in quant, readily ID the subjects being tested in all of quant and SC, applicable rules but CR remained an area of struggle (even though I had 83% accuracy on my most recent GMATprep). The final 3 weeks I was studying ~2-3hrs/day and an avg of ~16/week.

I would like some feedback as to when I should retake the test. Should I take a week off or attempt to carry this momentum until my my next possible test date in 4 weeks? I have exhausted all the MGMAT and GMATprep practice exams, how do I/should I utilize resetting these question pools? I have completed all the Quant easy/moderate questions in the OG13 and Quant2 and all the Verbal Moderate/Hard questions in the OG13 and Verbal 2- how do I get the most utility out of the questions I have already used? Any other insight to help me get to my score goal.

I've lowered my score goal to Q42V35, btw. The final date to achieve this score is July 8. Is it possible?

[StaceyKoprince]

I'm sorry that you had such a disappointing test experience.

We really need to figure out what happened and why it happened in order to figure out both what you need to do and whether you'll have enough time to do it by your deadline.

Both Q and V dropped, but V dropped more, making me wonder whether mental fatigue was one factor in the mix.

Admittedly, I put a little bit more focus on these sections on test day than I had on any practice exam

That extra focus on the first two sections would have caused more mental fatigue than normal later in the exam.

Your last two practice scores were pretty similar for Q but you had an enormous jump on V - and the second test was only 6 days later. What was the difference between the two tests?

I know at least one: the two free GMATPrep tests have identical IR sections - so, on the 2nd test, you wouldn't have had nearly the same level of mental fatigue from doing IR.

That would allow you to have the mental wherewithal to do really well when you got to verbal - but you wouldn't be able to repeat that outcome on real test day. If this is part of what happened, then your verbal score on that test was somewhat inflated. The good news is that you do have the intellectual capability to score that well on verbal, because you did so once - but you're going to have to train to be able to do so even after going through the full essay, IR, and quant gauntlet.

Here's what mental fatigue can feel like:
http://www.manhattangmat.com/blog/index ... you-crazy/

Also, take a look at this; were any of these issues for you?
https://www.manhattangmat.com/blog/inde ... t-wrong-2/

Tell me what you think about all of that so that we can figure out where to go from here.