Debrief: 750 in 1 month-->780 1 month later. Error analyses attached

• Would love to take a look at the error log you reference a few times. Sounds like careful use of that proved extremely valuable to you. Could you share this either on this thread of directly?
• As it relates to "drilling questions" from the OG, were you just creating question sets for yourself from the test bank in the OG online portal? Presumably you were setting the difficulty to "hard" only?
• When you were drilling your 15-20 question sets initially, were you mixing verbal and quant? Or would you alternate between the two topics for each set?
• When you talk about "error analysis", how did you think about dissecting your errors? Were you more focused on recognizing a topic that gave you trouble, a strategy that you found useful to solve a specific kind of problem, understanding why the right answer is the right answer, something else? I'm not sure I am as efficient as I could be when it comes to my own error analysis, so trying to understand your approach.
• As it relates to timing on the actual exam, were you finishing with plenty of time to spare, or were you using the whole time that is allotted?

• As it relates to "drilling questions" from the OG, were you just creating question sets for yourself from the test bank in the OG online portal? Presumably you were setting the difficulty to "hard" only?
  yes, I used the OG online portal set to level 700 only. But I found GMATClub & Manhattan Prep's question banks way better for Quant practice.
• When you were drilling your 15-20 question sets initially, were you mixing verbal and quant? Or would you alternate between the two topics for each set?
  I always found doing them separately to be better to learn--if I mixed it up it was harder to remember all the errors I was making, so I tried to focus 1 whole day for verbal sets, then 1 whole day for quant, etc.
• When you talk about "error analysis", how did you think about dissecting your errors? Were you more focused on recognizing a topic that gave you trouble, a strategy that you found useful to solve a specific kind of problem, understanding why the right answer is the right answer, something else? I'm not sure I am as efficient as I could be when it comes to my own error analysis, so trying to understand your approach.
  I really felt like this was the part that helped me get better fastest--I think I was good at correctly classifying and identifying the types of errors I was making.
  For example: at the start it was obvious almost all my errors were just being bad at DS, so I focused on getting over common DS traps first; after that I got past that became clear I was making consistent errors on number properties/algebra/geometry PS questions, so I started classifying errors by those categories next; and finally the last types of errors I kept making had to do with really specific number properties errors (not checking assumptions for even/odd #s, pos/neg #s, not checking of "x=0" was a possible solution, etc). The error analysis really had to adapt over time as I got better to target the incremental errors I was still making at each stage.
• As it relates to timing on the actual exam, were you finishing with plenty of time to spare, or were you using the whole time that is allotted?
  always had at least 5 min left for Verbal--that was much easier for me. For Quant I never had any time left; always felt like I had to be super focused the whole time to finish all the questions. But that's me--I'm weaker on Quant and had to spend more time on it.

• What specifically are you referring to when you talk about the Manhattan Prep question banks? As far as I can tell, they have the PDF with problem sets from the OG, and 25 questions per topic in what they call a "Question Bank" section (which I believe are the same questions every time you take it). Are you just referring to one of these, or is there some section of their site I haven't found?
• How many question sets per day do you estimate you averaged? Saw you mentioned you could do ~1-2 hours at lunch and presumably a few hours after work. By my math a 20 question set would likely take ~40 minutes + 40 minutes for review would put you at ~1.5 hrs per set - would you say you did like 2-3 sets per day?

Attachments

• Example of a practice exam log I used to track the types of errors made per exam
• Example of "cheat sheet" notes for Verbal and Quant I kept for myself to reflect the stuff I personally needed reminding of based on my errors--this won't be the most helpful for everyone but reflects more my personal weak areas
• Example of error logs for Verbal and Quant in the 2 weeks between my two tests

My Advice

• What helped me improve my score in a short period of time was consistent, repeat drills; being able to quickly understand and classify the types of errors I made so that I cut down on my error rate quickly[/b]; and recognizing that there is ultimately a limited scope of knowledge that the GMAT covers, and a consistency and logic to the types of questions they ask.
• The approach which I found most effective for the GMAT was repeated sets of drills of 15-25 questions each, interchanged with error review.
• Error review took up 50% of my time. It was the most important part of practice because it was the only part that helped me cut down on future errors.
• The goal in doing drills was to make consistently fewer errors in each set moving forward.
  
  • If you're still making the same % of errors after 3-4 sets, I'd go back to your error log and re-analyze it. If I was still making the same % of errors I was wasting my time. To me questions were only valuable if they exposed me to new types of questions I hadn't encountered before that were an opportunity to reduce errors.
• I spent 50% of my time on error review; maybe you'll be faster or slower than that, I don't know.

• I thought of studying kind of like having a giant map of a region that's hidden at first. It's your goal to slowly reveal different bits of the map, piece by piece, until it's fully visible. Every test has a limited scope of content. It might be a lot, but it's finite. Your task is to figure out how to tackle that body of content in a structured and efficient way.
• Score improvement is about focusing on your weakest areas, so you won't be effective if you consistently shrug off practice questions you get wrong by making one of two excuses:
  
  • "That was too hard. I'm probably not going to see that on the real exam, so I'm just going to ignore it."
  • "That answer/question doesn't make any sense. This exam is unfair. I disagree with the answer given."
  As a general rule, I'd say any question you see on a practice exam from GMAT Prep/Manhattan/Veritas is worth your time to study and understand. It will pay off on exam day. The GMAT is a fair and logically structured exam, covering a limited scope of content. Your score reflects how much of that content you have taken the time to understand. If you accept full responsibility for your mistakes you will be willing to spend the time and effort to correct them instead of blowing the hard ones off. Do not make excuses for your errors.
• I wouldn't be down on yourself if you don't see steady improvement. Some of the CAT exams are purposely harder than the real thing so you should expect lower scores on those. Despite my high score I was pretty hard on myself about Quant the 2-3 weeks before the exam and kind of thought "maybe I'm just not as good at this as other people." In the end I just kept practicing and proved myself wrong.

• GMATClub question banks were excellent and key in increasing my score from 750 to 780.
• Go through GMATClub's DS and PS question directories by level. I went through all 700-level categories and did 4-5 of the questions that were most bookmarked from each category.
• Number properties are the most important section to master for quant. After that, statistics/sets, geometry, and algebra/inequalities are probably second most important. I did separate drills for each category on GMATClub.

• Read PowerScore CR Bible's chapter 1 and skim the rest to understand the basic question types.
  
  • You should be able to map out every question prompt correctly into premises and conclusions. No matter how simple the prompt, I wrote out by hand an outline summary of every prompt in premises and conclusions.
  • Know the common traps. PowerScore CR Bible does a good explanation of some.
• In viewing the premises and conclusions, you should be able to identify the key logical factors in the argument: the objects/qualities/actions that lead to certain consequences.
• Language also matters: pay attention to probability/descriptive words like "most likely", "always", "never", "frequently", "might", etc. On top of the structure of the logical relationship described, these add nuance of the strength of the relationship.
  
  • E.g. "Snowfall causes traffic jams", "Heavy snowfall always causes traffic jams", "light snowfall may cause traffic jams", etc.
• In some ways, CR is the most math-like part of Verbal. The logic is not that different from logic in math, so there is no reason why someone who scores high in Quant should be weak in CR given enough practice. Logic in CR is simply expressed in adjectives and words, whereas logic in math is expressed in symbols.

• I spent 3-4 min reading each passage before going to each questions.
• Know the common traps. My most common error was narrowing it down to 2 answer choices, then going for the one that sounded superficially more similar to the text rather than the one that was logically the stronger relationship. Don't be tricked into choosing an answer that is phrased to sound similar to the passage when it logically is not the best answer.
• A lot of the passages I read were on topics I had no background in (black holes, microbial evolution, etc). That's fine. Jargon does make the reading slower to plod through, but it ultimately should not hinder your understanding of the logical structure of the passage.
• If a question addresses a specific sentence, always go back to the sentence to re-read the sentences around it and understand the context.
• Probably more so than CR, language and tone matter in RC. It communicates subtle nuances about the author's attitude and the strength of their attitude. These reading comprehension skills I admit are hard to acquire/improve in a short period of time. All I can recommend on that front is to just keep practicing.
  
  • I do not recommend spending much time reading The Economist, etc. in that regard. The best practice you can get is from actually doing questions. Doing questions forces you to engage with the text instead whereas much outside reading is done passively. When you practice, you want to be forced to think critically about the passage in a way that is hard to do on your own without structured prompts.

Debrief

• Started studying: July 2018.
• 1st exam, Aug 10: 750 (Verbal 46, Quant 48) / AWA 6.0 / IR 8.0
• 2nd exam, Aug 29: 780 (Verbal 48, Quant 50) / AWA TBD / IR 8.0
• My experience will be helpful to you primarily if you are already testing in the 630-700+ range on Manhattan / Veritas CATs and in the 700 - 800 range on GMAT Prep.
• The 30 pts improvement from my already high score in 2.5 weeks was due entirely to targeting my weaker areas in Quant. So my experience will be somewhat more relevant to those people who have already mastered Verbal and looking for tips on closing the final 30-50 point gap based on improved Quant performance.

• 7/01 - OG Diagnostic Test: 680
• 7/11 - MGMAT: 670 (40 verbal / 41 quant)
• 7/23 GMAT Prep: 720 (42 verbal / 47 quant)
• 7/23 GMAT Prep: 770 (47 verbal / 49 quant)
• 7/31 - MGMAT: 44 verbal (no full score; accidentally skipped quant)
• 8/02 - MGMAT: 650 (40 verbal / 39 quant)
• 8/03 GMAT Prep: 770 (47 verbal / 49 quant)
• 8/06 GMAT Prep: 770 (44 verbal / 50 quant)
• 8/07 GMAT Prep: 780 (50 verbal / 50 quant)
• 8/08 GMAT Prep: 50 quant (no full score; accidentally skipped verbal)
• 8/09 GMAT Prep: 770 (47 verbal / 50 quant)
• 8/22 Veritas: 650
• 8/25 Veritas: 670
• 8/26 - MGMAT: 650 (36 verbal / 43 quant)
• 8/27 - MGMAT: 710 (45 verbal / 42 quant)

• I am a native English speaker.
• A relatively balanced academic/professional background between math/verbal.
• Good work hours: despite working in finance, I worked a very manageable 9-to-5 workweek for the entire 2 months I studied. I also always had time to practice questions for 1-2 hours during lunch regularly.
• A good study environment: I live in NYC, which is full of good cafes to study in. For the past two months, I worked in a dedicated co-working space in the neighborhood to ensure 0 distractions.

• I knew 750 was a very good score.
• But in my gut I had expected to do better based on my GMAT Prep results of 780. That was the number I had in my head.
• I felt that I would have attained that 780 had I known to study more of the right content for Quant. The questions I got wrong didn't reflect my actual ability but rather the fact that I had given more weight in studying to the wrong topics.
• My ESR report made up my mind. In quant, it showed 43% of my answers were wrong in the 3rd set of questions and 57% were wrong in the last set. I got only 33% of geometry questions and 50% of inequality/algebra questions right. I know I can do better than that.
  
  • As others here have observed, it was clear to me that I spent too much time going through advanced combinatorics/probability drills when the heart of GMAT quant is number properties, geometry, and algebra.

• Quant drills. Divided my time evenly between GMAT Club question banks and Manhattan, Veritas, and GMAT Club CATs. The purpose was to expose myself to distinctly harder Quant questions than the GMAT Prep tests and the Official Guide Test Bank, and harder questions than what I'd encounter on the GMAT exam itself.
  Drilled these every day, focusing on geometry, number properties, and algebra; and to a lesser degree on sets/statistics.
  Veritas Prep, 650
  MGMAT 8/26 650 (Verbal 36 / Quant 43)
  MGMAT 8/27 710 (Verbal 45 / Quant 42)
• Spent minimal time on verbal. I was confident I had it down already. Returned to verbal practice once every 3 days to make sure I wasn't backsliding but that was enough.
• I ignored IR completely which is easy relative to Quant/Verbal. I did make sure to test that my IR ability had not receded with a question set on the Official Guide Test Bank the day before the exam.
• Aug 29, Exam day. opposite of the first exam experience--after drilling through hundreds of GMAT Club questions, I found Quant noticeably easy compared to my practice experience rather than difficult. Verbal I found as straighforward as before.