Hi
I have already posted quite a bit on this subject with various different angles of discussion. Today, the discussion is more about :
1. "How Much" and "if" - difficulty level says anything about the actual scores and performance in a test
First, what I have read and heard is that difficulty level can be an indirect way of measuring how one is progressing on the test. It can also be used for practiced test takers to actually modulate the path the exam is taking - basically guess and bring down the level then take it up and bring it down a notch - so that within humane capacities one can achieve a target score which in my case is 700 +
This however does not seem to hold in my case. I will come more on this after discussing the second.
2. The truth behind "GMAT prep" official exam practice tests are the actual estimation of your ability as far as the exam taking skills as depicted by scores goes.
As far as practice tests goes by level of difficulty and true estimation of what will be the score in the final exam this is what I have read :
Veritas/Manhattan - On the higher end of Difficulty
Kaplan - Medium to Hard - but might actually be most real to the actual test experience given the interface and types of questions etc
GMAT Exam prep - gives a true estimation +/- 20 at the test.
Now, the above is true mostly. While practicing I have found that indeed Veritas is the most difficult with Kaplan and GMAT exam prep of similar difficulty. However, in reality I have found that the questions in the actual exam though not difficult are full of calculations [if given 3 mins/q I would have a 100% score], unlike the preparatory courses which focuses on enhancing your logic and abilities of discerning and solving a problem within 2 mins. Now this poses a problem - as the first 10 are important, so no matter what time and accuracy is a must[keep cross-checking yourself all the time]. This holds true in both Quants and Verbal. But then you cannot completely rest in the later half and make decisions of which to guess etc as you go along. For Q one can still decide given its like 1 problem and you make a choice if it can be solved correctly in 2 mins or better to just guess and move. I have had this strategy backfire sometimes - though always it seems unfortunately only at the actual. Now that's a problem then and there is no meaning of any practice tests at all...
Now in verbal, however, it is difficult to do the above as if you are stuck with a long complicated passage right around the borders Q 8-12 or Q 22-25 and one wants to put in only minimal effort and guess and move on, what I have found is that the scores show that in reduced numbers always. So then in that case the mindset I went in was I will put in full effort for all SC and all RC with a 50% effort in CR. That generally gave me a 60% -70% accuracy in verbal overall in my practice tests [with 50% on a bad day]. Off course, though the real GMAT seems to ask for super human abilities specifically at the test center. As none of the above works, so then I have to 90% upwards accuracy in everything - which would only then give me a 700 + score [that too doubtful?]. This is unlike anything I have read or heard from anyone with a 700 score.
GMAT is not a school paper exam that being accurate 90% will give a 90% score. Then the adaptive nature and being able to strategize and score according to target would be a ballgame for anyone, but it is not. There would be no requirement to take practice tests etc. One would memorize a list and like high school would vomit the knowledge and get a 750+. Yet multiple people have said that despite getting 10-12 Q incorrect in Quants and 8-10 Questions incorrect in verbal they have achieved 700+ scores. That is how they played their game. That is 58% correctly answered questions can push you upwards of 650 even into the 700 zone depending on the level of difficulty, which were correct [chain of questions vs alternate].
I am posting this to ask if all my assumptions are incorrect? It seems in my case it is, and I am getting scored according to accuracy - i.e 58% correct questions means 580 will be the score [according to ESR] and not actually 85% percent given its a 68 question test and not a 100 question test, and then according to the cohort a relative score is not obtained, but like a paper based CAT test a handwritten average is being done. I have multiple friends who have previously given the CAT for example and that is what their advice has been. Always be 95% correct - then only there is any chance of getting into the CAT. However, given GMAT is an US based exam and even though I have been a TA at a US school, and have relatively graded graduate level students myself in the past, my assumptions of any relative grading on the GMAT seems to be completely off-kilter.
In my experience any advice given on this forum or anywhere seems not to work for me. So, would anyone who has obtained 700+ scores multiple times [supposedly that is the requirement], and know the tricks of the trade be able to give me some advice. Specific advice on each Verbal Section would be awesome, and Quants overall macro - pacing.