Study and Strategy questions relating to the GMAT.
LuisK608
Course Students
 
Posts: 2
Joined: Wed Feb 19, 2020 7:03 am
 

CAT scores stay the same

by LuisK608 Tue Apr 21, 2020 11:21 am

Hi there,

I got really upset when I did my 4th CAT today and scored the same again. I want to take the GMAT mid of may. My goal is around 650.
Here is an overview of the CATs (always one of the 6 I got for free with the Manhattan Prep All the Quant and All the Verbal books) I took. I did the AWA only on CAT 1 (as it is now removed from the Online GMAT) and had real test conditions, did the test in a seperate room at the office and didnt pause etc.:

Date/Overall Score/Quant Score/Verbal Score/IR Score

CAT 1: 19th February: 480 / 29 / 28 / 4.5 (Without studying)
CAT 2: 20th March: 580 / 42 / 29 / 3.8
CAT 3: 08th April: 580 / 39 / 31 / 2.4 (After finishing "All the Quant" and "All the Verbal")
CAT 4: 21th April: 580 / 38 / 32 / 3.8 (After concentrating on my weaknesses from Bucket 2)


So I think I have 3 problems:

1. Quant: In CAT 4 I was ahead in time like 3/4 minutes until the last couple of questions. I have about 5/6 questions in Quant, where I have no idea how to approach the problem (especially Geometry cause I'm bad at it and put it in Bucket 3 for now), so usually I just bail quickly and guess. That means I have some more time on other problems, but often more time doesnt really help me to get the right answer. I think in general I'm just not good enough in math concepts and rules. Do you have any advice for me on which topic and which ressource to concentrate? The last 10 days, I only studied Algebra and FDPs and it didnt really change anything...

CAT 2, 3, and 4 Quant Summary:

Content Area / % correct / avg time correct / avg time incorrect /avg diff correct / avg diff incorrect

Geometry 18% 1:21 1:13 550 650
Number Properties 47% 1:54 2:00 580 680
Algebra 52% 1:55 2:05 570 640
Word Problems 75% 2:12 2:11 560 610
Fractions, Decimals, & Percents 50% 2:08 2:16 540 640


2. I finished both the All the Quant and All the Verbal book 10 days ago and am not sure what exactly I should do now until the GMAT exam in mid of may. I started to do an error log on OG2020 problems while doing the OG problems (and repeating the chapters in the books) recommended by ManhattanPrep. I always try to finish a problem in under 2 minutes and sometimes I review them directly and sometimes I do sets of problems (still trying to stay under 2 minutes). I feel like sometimes I should invest a little more (as I'm bailing on 5/6 problems anyway). What do you think about that?

3. In Verbal, its the same as in Quant. I think if a problem seems hard to me, I have to bail quickly, thats why in CAT 4 I was several (up tp 6 minutes) ahead. I guess next time I'm doing a CAT, I'll just try to do the harder problems anyway as long as I'm in time?


I also have a question regarding the GMAT Strategy Guide Question Banks. When should I do them and how are they part of the study plan?


I'm writing this in the afternoon after doing the 4th CAT and I'm thinking to take a GMAT Practice Exam from mba.com next week as I heard that especially the Quant section in ManhattanPrep CATs is much harder, as I really want to see where my score is currently.



If you need more info about my problems, I'm happy to go more into detail. I would be very happy to get some advice on how to continue studying until the real GMAT!


Best regards
Luis
StaceyKoprince
ManhattanGMAT Staff
 
Posts: 9348
Joined: Wed Oct 19, 2005 9:05 am
Location: Montreal
 

Re: CAT scores stay the same

by StaceyKoprince Sat Apr 25, 2020 12:58 pm

Hi! Yes, it's frustrating, but it's not uncommon for scores to stagnate a bit like this. Progress on tests typically isn't linear—you have to get to a certain point before things start coming together enough to see an actual lift in your score. I'll answer your questions mostly in order.

I do see that your V score has been edging up a bit, but your Q score has been dropping. Any theories as to why? If we could combine your highest Q and highest V score, you'd be at a 610. (I know you want more than that, but at least it's something.)

Next, have you used our CAT Review Tracker spreadsheet to analyze your results? You talk about Bucket 3, so I know you've either read my article or used the spreadsheet (or maybe both?). If you haven't followed the detailed steps in the spreadsheet, I'm going to suggest that you do so (warning: it takes hours), because that level of analysis is what allows you to figure out precisely where you have the best opportunities to get better and how.

For example, you classify all of the problems into one of several categories. The different categories are going to lead you to different ways to review / get better at that problem. If you made a careless mistake, the issue is that you have to build some new good habit to execute on whatever the mistake was (and, usually, you have to break a bad habit that left you open to that type of careless mistake in the first place).

It will also help you to see precisely where you made good timing decisions and where you didn't. You mention that you bail on some problems (good!)—are you bailing quickly enough (within 30 seconds)? If so, good. If it's taking you closer to a minute to decide, bail faster. You also mention that you use the extra time to invest on some problems but that doesn't really help. That tells me that (a) you're still choosing to invest extra time on problems you don't really know how to do and therefore (b) the task here is to learn how to make better choices about where to invest that time.

For example, careless mistakes are often due to rushing a little bit / just not taking that extra 5-20 seconds to write out your work or do an extra step or test one more case.

Another category we use when you classify your questions is "wrong but I get it now." What is is that you didn't know before but you do know now? And/or what do you need to practice / get better at? Or what didn't you notice that you realize now? And most important of all: How are you going to recognize this thing the next time you see it on the exam? Literally, you need to get to the point where you can say "When I see <a certain characteristic>, I'm going to <take this specific action or approach>."

We call this "Know the Code."
https://www.manhattanprep.com/gmat/blog/the-second-level-of-learning-to-take-the-gmat/

Go ahead and read that right now, then come back here. Later, you can also search the blog for Know The Code to find other posts that talk about this concept and show how to apply it to specific problems.

If you have gone through the full CAT Review Tracker spreadsheet, give me some more detail about what came out of the analysis that's listed there. If you haven't, go ahead and do the analysis. (In Atlas, you'll find a link to the tracker and a link to the videos that explain how to use it in any section that assigns a practice test.)

Another thing I want to know by the way: You seem to suggest that your timing is ok, but is it possible that you are exhibiting Up and Down timing? The spreadsheet will tell you what this is and how to tell whether you're doing it, but basically it's "I spend too long on some but I also rush on others, so I'm on time overall throughout the section...but I'm still not making the best decisions I could make about where to spend my time on specific problems." From what you've said, it sounds like it's *possible* that this might be true for you.

I always try to finish a problem in under 2 minutes and sometimes I review them directly and sometimes I do sets of problems (still trying to stay under 2 minutes).


You need to *average* 2 minutes, not be under 2 minutes. Some problems will take you 1:30, but others will take 2:30—and that's fine. My own personal "this is too long" time is 1+ minutes beyond that problem's average time (so, on a quant problem, 2m is the average and 3m is too long). The idea there is just that, if it really takes me THAT long to do the problem, then there's something I'm missing or not getting—and, if that's the case, then I'm a lot more likely to get the problem wrong even if I do spend 3m on it.

If I get a problem that's a harder one among the mix that I'm earning, but it's still something that I know how to do, then I'll likely need 2:15 or 2:30 or even 2:45 to do it. But I'm not tossing extra time at it to see whether I can figure it out with more time. Rather, I realize in the first minute or so that I do have a good plan to approach the problem, but that plan is just going to take somewhat longer than the average. And I'm willing to invest that time because this problem is something that I can often do.

I guess next time I'm doing a CAT, I'll just try to do the harder problems anyway as long as I'm in time?


Doing the deep analysis with the CAT Tracker spreadsheet will answer this question for you. When you review problems afterwards, you'll see which ones are the ones that you actually could have gotten right if you'd invested a little extra time vs. which ones would have been too hard no matter what. And you'll ask yourself WHY / HOW you could recognize that next time (know the code). So on your next test, you'll be better prepared to make good decisions about when to invest and when not to invest.

Basically, when you review your test (or any practice problem), you're not just reviewing and analyzing how to do that individual problem. You're also reviewing and analyzing your decision-making (on time management, on the process / approach you used to solve, on how you decided whether to bail). The GMAT is a test of your Executive Reasoning skills at heart—and you actually have to study that, just like you study the underlying content.

The strategy guide question banks are just extra problems—they're not assigned in Atlas. A lot of people never do them, but if you do want any extra practice with any of those topics, they're available to you.

I do think that the Quant section of our CAT can be harder than the real thing. It is especially harder (on purpose) for people who are struggling with the following:
– Time management
– Trying to solve the textbook way vs. using test-taking strategies / looking for the fastest and easiest way
– (Tying the above two bullets together) Overall decision making about where to invest and where not to invest, as well as what approaches to use when you do solve

Having the above issues makes it more likely that your score stagnates and also makes it more likely that you experience score fluctuations—which can mean a sudden, big drop on a practice test or the real test. So yes, please feel free to try an mba.com test. If you do see a big difference, know that this means you're likely struggling with one or more of the above items. And that means that you could be okay on the real test (because sometimes all of the pieces do come together well), but your timing, decision-making, and/or solution approaches can lead to variability in performance that can result in a score drop.
Stacey Koprince
Instructor
Director, Content & Curriculum
ManhattanPrep
LuisK608
Course Students
 
Posts: 2
Joined: Wed Feb 19, 2020 7:03 am
 

Re: CAT scores stay the same

by LuisK608 Mon May 04, 2020 10:07 am

StaceyKoprince Wrote:Hi! Yes, it's frustrating, but it's not uncommon for scores to stagnate a bit like this.


Hi Stacey, thanks a lot for your reply!
I am going to answer your questions:

1. CAT Review Tracker: Yes, I’m using the tracker and spent quiet some time on reviewing everything, but your hints helped me to analyse my CATs better. I saw that I bailed too late on 3 Q problems (took around 1 minute), In V I’m guessing faster so that’s ok.

2. To go more into detail about the CAT Review Tracker (referring CAT 4):

a. Quant:
- CAT 3: All 300-500 and 500-600 right, but sometimes >2 mins, need to be a bit quicker there
CAT 4: 1/3 of 500-600 wrong!! I got less 600-700 questions
I think I need to get better at the basics (means easy questions and basic math concepts, I know this is a big weakness, I’m already studying easy multiplication, addition and so on) to get most of the 500-600 problems right, cause this will get me to a score around 600 in Q
- bailed on problems i knew i cant solve cause i didnt have an approach
--> Dont take too much time on the next problem immediately after you bail one, save the time wisely and dont waste it on easy questions
when you bail, the difficulty level probably decreases so dont waste time then
- When I find myself more than a few minutes AHEAD, I will:
be careful where and when to invest the time
- When I find myself more than a few minutes BEHIND, I will:
choose a good problem to bail and guess immediately, not bailing on a problem i can solve, so read problem quickly and decide to bailor solve it

- Bucket 1: Word Problems (75% correct)
- Bucket 2: Question Type(s): exponents&roots, fractions, percents, ratios
- Bucket 3: Geometry (25% correct), I try to bail on all Geometry problems as I studied this topic only once

- What did you notice about your timing from this different view?
faster on PS (average correct problems 2:00 on PS and 2:18 on DS) but more correct answers on DS (71% to 35% on PS)

- How does timing impact whether you get a question right or wrong?
spent more time on correct answers (2:15 min) and less time on wrong answers (1:50 mins)


3. You seem to suggest that your timing is ok, but is it possible that you are exhibiting Up and Down timing?

In Q I bailed on Problem 5 and 7, so I was 2 minutes ahead, I used this time on the next 7 problems, so I was at 0 in my time balance at problem 14. I bailed on problem 15 and 16 fast and was 4 minutes ahead. Then I used 3 minutes on a 300-500 question (problem 17) and another 3 minutes on a 500-600 question (problem 19). I continued to bail on 1 more question (problem 25) and used this time on the last problems.
I guess I have to get better in deciding when to use my extra time, cause I feel like my guessing strategy is fine (I just need to bail faster). But I have difficulties to see when I could use some more time to get the right answer...and I dont have any idea how to get better there...

In V I got ahead quick, after problem 3 I was 1 minutes ahead already. My goal was to bail faster here (especially when I see long and complicated SC problems). Problem 11 was this kind of question (700-800 CR), I bailed in 20 secs and it was a lucky correct guess. After that I was 3 minutes ahead. I got A 700-800 RC passage on problem 13 and continued to go fast (as I didnt understand the passage at all). So after problem 15 I was 4:30 ahead. I still continued to go as fast as possible while still getting correct answers and not to do careless mistakes. The next RC passage was easy so I was 5 minutes ahead afterwards (problem 20). On problem 24 and 26 I got 700-800 SC questions which I bailed fast so I was 6:30 ahead.
-->Here I think I could have tried to solve them as I was ahead so much. And I got all except one easier question right in between.
Like at problem 31 around that I noticed how much I’m ahead and that I have just a couple of questions left so I spent more tome on the last questions. I still contiuned with the strategy that if your down to 2, just compare once and decide.


4. When you review problems afterwards, you'll see which ones are the ones that you actually could have gotten right if you'd invested a little extra time vs. which ones would have been too hard no matter what. And you'll ask yourself WHY / HOW you could recognize that next time (know the code).

I think this is a big point for me so I’m going to work on that when reviewing the OG and CAT problems.I still have difficulties to categorize problems and to get the connection to previous experience and how I can recognize similiar problems in the future, and often I just have no clue how to approach them. I think thats just a learning process...

I’m going to do an original GMAT test on mba.com tomorrow to see how my score is there and then continue to work on my weaknesses. I feel like I can get more points in V when I am able to solve some of the 700-800 problems, and that’s just a timing issue currently.

Thank you so much for your help already! Lets get this! Really helped to motivate myself :)
StaceyKoprince
ManhattanGMAT Staff
 
Posts: 9348
Joined: Wed Oct 19, 2005 9:05 am
Location: Montreal
 

Re: CAT scores stay the same

by StaceyKoprince Mon May 04, 2020 7:25 pm

Hi! I think your first attempts to post got caught in the spam filter because it looks like you might have tried to attach images? And then after that, I think it put your later attempts in the spam filter because you were posting really long posts in a very short period of time. :)

FYI for future: Don't include images. URLs are also likely to get you kicked to the spam folder. Really long posts can also trigger it, so one thing you can try is splitting long posts into 2 or 3 parts. (I'd still try first to post it as one long post.)

Yes, if you miss too many lower-level problems you'll never get yourself up into the higher scoring levels. If you do have holes in your foundations, you can sometimes get lucky depending on the specific mix of questions you get—but if you get unlucky and get more lower-level problems in your areas of weakness, that can really hurt. So it's important to make sure that you have a really solid foundation across the lower-level material.

Dont take too much time on the next problem immediately after you bail one, save the time wisely and dont waste it on easy questions
when you bail, the difficulty level probably decreases so dont waste time then


I agree with the first half but not the second—and for the same reason. :D The basic principle: Don't decide what to do on the next problem based on what happened on the last problem. Decide what to do with each new problem based on that problem's merits. (When you bail, you still have a 20% chance of getting lucky...so yes the odds are better that you're about to get an easier-according-to-the-database* problem, but you never know.)

*Also: The difficulty level assigned in the database is based on the performance of the ENTIRE population of test-takers. Your strengths and weaknesses are different than the average across the entire population. So even though the next question might be easier-according-to-the-database, it might be harder for you—or vice versa.

So back to the mantra: Each question gets its own look. You decide what to do based on the details of that problem. If it's something you hate / are terrible at, you bail as soon as you recognize that—even if you are on time or ahead of time, even if you also bailed on the last problem. I once bailed on the last two problems in a row on an official Q section even though I still had 5 minutes left and could have tried them both. (One was a cylinders problem, one was combinatorics. I hate both! And I still had Verbal to come and didn't want to waste the mental energy on problems I hated. They suck up extra mental energy proportionally. No thanks.)

Let's tweak some more things:
- When I find myself more than a few minutes AHEAD, I will:
be careful where and when to invest the time


Not quite. You aren't choosing to invest more time just because you have it. That would be like, "I need $1 million for my new business and you haven't invested any of your new VC fund yet, so you'll fund me, right?" Not unless you can prove you have a great business opportunity.

If you find yourself with more than 2-3 minutes ahead, slow down in general. You're likely losing points to careless mistakes because you're rushing. Make sure you're writing out your work. Read more slowly / carefully. That sort of thing.

- When I find myself more than a few minutes BEHIND, I will:
choose a good problem to bail and guess immediately, not bailing on a problem i can solve, so read problem quickly and decide to bailor solve it


Yes. But... I need to make sure of one thing. A lot of people interpret the above as "Don't bail UNTIL I'm behind on time." That's NOT a good plan. You bail when a problem is terrible for you. That can happen any time, from the very first problem to the very last. That can happen when you are already 2 minutes ahead on time. It doesn't matter. All that matters is that you know this problem is terrible. Spending any time, even extra time, isn't going to make it any better.

If you are actually following that mantra as you go, then it's unlikely that you'll find yourself more than a few minutes behind. :)

Buckets
I think it's fine to put most of Geo into bucket 3, but you do need to make sure you know the foundations. If you get a couple of easy Geo problems, you don't want to let them go just because you didn't memorize the formula for the area of a triangle (and that's all the problem asks you to do). That's going to bring your score down.

So do some work here to identify the kinds of Geo problems you can and can't handle (start from the beginning of a quant chapter in OG and see how far you can go). Know the rules / formulas / basics and how to recognize when the problem isn't too annoying to handle—and also how to recognize when the problem is starting to get into "bail" territory for you.

Where does the rest of Alg fall in your buckets? What about Number Properties?

I guess I have to get better in deciding when to use my extra time, cause I feel like my guessing strategy is fine (I just need to bail faster). But I have difficulties to see when I could use some more time to get the right answer...and I dont have any idea how to get better there...


When you retry the bail problems now, how do they feel? Do you still think those were the right ones to bail on or, now that you've tried them, do you wish you'd actually done some of them on the real test? What characteristics made you initially think to bail? And what characteristics made the problems actually not that bad / doable?

Are there other problems you did try but you then realized they should have been bail problems? What characteristics made you think they were ok at first and then why did they actually turn out not to be ok / you should have bailed?

That level of analysis is what allows you to make better decisions next time. Keep doing that analysis and refining your decision-making.

Love your last line! :cool:
Stacey Koprince
Instructor
Director, Content & Curriculum
ManhattanPrep
BlairD379
Course Students
 
Posts: 2
Joined: Sat Feb 08, 2020 4:57 pm
 

Re: CAT scores stay the same

by BlairD379 Tue May 19, 2020 1:27 pm

Following up on this, I have been stuck at a 580 too coincidentally for the past month.

In pulling up my latest diag (CAt 3), its clear in Math i am scoring well on 500-600, but I am absolutely shanking 600-700 problem questions which is why I am stuck at 580. To be clear Im targeting a 700 so i realize i have a ways to go but I think this is quite fixable seeing as a went 2/13 on 600-700 problems and i saw 4+ 700+ problems.

On this latest diag with verbal from questions 14-25 i got 2 problems right in an 11 problem stretch (seems like fatigue to me). Many of these problems were reading comprehension and critical reasoning which i usually do better on per my last CATs. I spent a ton of time on grammar since last cat which saw a marked uptick so I'm actually less worried about my scoring on verbal as it actually was due to a stronger performance on grammar and weighted down by a poor reading and cr which i didn't spend a ton of time on.

In short, the easiest fix appears to be getting alot more of these 600-700 problems right. Is there a way to only do these types of problems by difficulty?

Thanks
StaceyKoprince
ManhattanGMAT Staff
 
Posts: 9348
Joined: Wed Oct 19, 2005 9:05 am
Location: Montreal
 

Re: CAT scores stay the same

by StaceyKoprince Fri May 22, 2020 9:48 pm

Hi, Blair! I'm going to give you a quick reply here, but could you start a new thread for us to discuss your "case"? Otherwise, I'm going to mix up data from the two of you and then I might mix up my recommendations. You can copy-paste what we've already discussed into the new thread and then we'll go from there.

I understand why you asked what you did, but you actually don't want to do only 600-700 level problems. :)

In order to lift your score, you need to concentrate on 3 things; your question addresses only the first one:
(1) How to get better at answering the ones that you are close to getting right but can't quite get right yet.
(2) How to get *faster* at the ones that you can already answer—but they're at the upper end of the ones that you can answer correctly right now.
(3) How to recognize the problems that are just too hard for you so that you can bail faster.

Basically, you're always going to see a range of problems. Let's imagine that you see problems in the 500 to 700 range and you answer:
– The 500 to 540 ones correctly and fast
– The 550 to 580 ones correctly but in normal time or maybe a little extra time (that you picked up from answering the 500 to 540s fast)
– The 590 to 640 ones incorrectly but you were able to do some of the work properly
– The 650 to 700 ones wrong (regardless of how much time you spend)

If you just focus on the 590 to 640 ones and you learn how to get some or all of them right, your question mix is going to change. Next time, you're going to have a lot fewer of the 500 to 540 ones—the ones that you can currently answer fast. So if you don't ALSO spend time learning how to get faster at the 550 to 580 ones, then you aren't going to be able to spend a little extra time on the 590 to 640 ones—and you will sometimes need extra time on those.

Basically, as you lift your score, the problems that used to be in your "middle" range now become part of your "easy" range—so you need those problems to continue to be in your problem sets so that you can keep working on getting more and more efficient with them.

And then, for the ones that are just too hard to do—you still want them in your problem sets because you have to practice being able to spot quickly (< 30 seconds) that this problem is just too hard and you have to practice having the discipline to guess randomly and move on when you do see one of these.

Finally, your skill level is not perfectly "level" across all content areas—you have strengths and weaknesses, as we all do. So there are some lower-level areas where you do actually need to work more on content, not just efficiency.

Question for you. On those problems that you did miss...why did you miss each one? Precisely? That kind of analysis will allow you to figure out what you need to do to get better at whatever it is. It could be that you didn't know the content, but it also could have been a careles mistake. Or you knew the content but messed up the process. Or you could have used a test-taking strategy (estimation, work backwards, etc) but you didn't notice. Or you did notice and did try but messed up the execution of the strategy. Or you missed this one and you *should* have missed this one because it was really hard—but you spent 2 minutes missing it when it was so hard that you could have recognized to bail on it 30 seconds in.

Go to that level of analysis with every problem. That'll tell you what you need to go study or practice in order to get some of those (or similar ones) right next time. :cool:
Stacey Koprince
Instructor
Director, Content & Curriculum
ManhattanPrep