Just back from Test centre, disappointed, something to share

Good to know that!! Great you will definitely help others and also crack it next time.

Keep it up!!!

Waiting eagrly to hear the reasons.. Almost similar happened with me. All I can think of is that probably i did not do well on SC. During practise I was always confident about CR and RC. In most of the practice tests my accuracy in CR and RC was also very good(from 70% to 90%). But for SC I was not very confident.

Hi all,

Here is my theory about my sudden and significant drop in verbal score in the real test.

Not really my theory. Someone talked briefly about this before.

I am so tired, so I will just type with the flow of my thinking. Hope the grammar errors will not bother you too much : )

Because in my real test, the questions were so easy, I am not a careless test taker, and I can identify many tricks. So I don't think the drop is because the real test are more difficult.

Some people said there was bug or GMAC's policy is to give a low score to first time test taker. I do not believe in that!

The main reason is the uncounted questions. We all know there are 10 questions in verbal are uncounted.

THink about that, if I consistently got 80% (32/41) correct in practise test. I will score late 30 (or early 40 maybe!?) in verbal score. So i become very confident that at least I can get mid 30 in the real test.

But what if among the 32 I answered correctly includes the 10 uncounted questions? then my hit rate will become 22/(41-10) = 70%. the drop from 80% to 70% means alot among serious test takers aim to score high in verbal and can bring you down to verbal score 30-32.

In prastise test, you do not have this problem, because all the questions count. This is the first effect made by the uncounted questions.

Even if GMAC erase all the uncounted questions, and made the test with 31 questions only. THe risk for drop in score is still higher than doing practise test with 41 questions. Think about if you have only 10 questions to the extreme.

Someone may argue why no significant in Quan. Because the hit rate for quan is normally high. For ppl aim at 50, 51, there will be less than 5 questions wrong. So the risk is much lower.

If you still understand what I am saying. Here are my recommendation.
- don;t rely on luck.
- assume the uncounted question will be the question you got right
- and all your wrong answers will be counted.
- aim at 95% high rate in practise test. If you cannot do that, you are relying on luck to help you score high.

Seriously, I don't think I did any of the CR wrong!

RC at most are wrong.

SC, i think I get at least 70%. half of them are very obvious SC, consider XY, compare apple to apple, tense ...

Oh! the magical "27" on Verbal, sorry on the score! I'm sure you'll do well the next time around.

IMO - You do have uncounted questions in both Quant & Verbal, and I feel its 7 & 11 respectively.
Sometime back, on GMAT Preps, I received :

V29 (20 Incorrect)
V29 (17 Incorrect)
...on 2 different test.

So, I'm guessing Prep too follows uncounted questions (I may be wrong)...did anybody else come across such instances? or notice anything similar?

-Pawan

That might be the case, just to add, I want to say, suppose if the test taker is particulary weak/not efficient in the particular stream of questions such SC or RC. And in the begining test taker gets those kind of questions more(than rest of the test), so the computer might put you in the inappropriate bracket of verbal score and simulate rest of the test accordingly. So the bottomline is follow the walker's strategy consider each question monster, that you have to kill else monster will kill you. And try to attain almost similar accuracy in all stream of questions.

I was the victim of that.

I tried to do all questions perfectly and even recheck every question.

I think even i have to guessed the last 2 questions will not be too harmful (i remember someone'e debreif saying that he guess the last 2 questions and still got above 40 in verbal).

Probably, the last 3 I guessed are counted questions

Looking at the V27 again, I am really paranoid about my coming GMAT test. If people with so good verbal ability can be floored due to whatever reasons (software bug, bad luck on doing all the experimental questions right and real questions wrong) then one can not rely on GMATprep at all.

Believe it or not, I tried mathematically what judokan is saying and apparently found correctly. Please see the attached xlsheet. What I did, I calculated the the first score with 11 uncounted questions and second score with all counted questions, I found that 41 on practice test can turn to 34 in the real gmat, if we account 11 uncounted questions. So my analysis reveals that a testtaker is more vonurable to this counted/uncounted gimmicks if he/she is between late 30 and early 40 verbal score range. below this and after this range, surprise is less. Remember, its just my mathematical model to represent the theory that we developed on this forum.

After doing this analysis all I can say-
Maintain accuracy level among all stages in the test, and for all streams of questions.
Try to gain mid 40s in the practice test so you can beat the element of surprize.

judokan .. it gets really scary for an average test-taker like me on seeing such a huge score drop for a high scorer like you ... But anyway I am sure next time you will hit it spot on ... and regarding your theory of the counted/uncounted questions ... to me, it makes sense ..... Truly speaking i was wondering recently that what will happen if i get the experimental questions right and the real questions wrong ... SCARY !!

All the best for your retake (If you are considering one)

Hey subarao,

Think positively, there are a lot people who are not affected or even benefited by this reason.

Again, think positively, this post should not discourage you. But it tell you something before the test so that you can better prepare.

My main fault is that I over invested much on 90% of the questions. And turned out that some of them were not counted. If I had a choice to do it again, I woudl have evenly distributed my time over the 41 questions.

And DON"T relax until you finish the question no. 41.

You will be fine. We wait for your good news.

Think positive mate. You can never guess score, untill you saw. Stay focused and stay confident.

Tks amitdgr, sure I will give another shot.

I hope my post can help people realize the difference between the practise test and the real one, not bcause of the difficulty, but of the experimental questions.

Some people said above that the experimental rules also apply to GMATPrep. I doubt that because if there are experimental questions, our score would not be so consistent. Also, what is the point adding experimental questions in practice?

Yes, do your best in every question, and the final score will suprise you.

Besides, GMATPREP software uses static question bank and real gmat uses dynamic question bank. So more people score higher score, tougher GMAT gets.. And even the experimental questions are there in the GMATPrep, it will not simulate the Real exam pressure. I guess the best way to use OG11 and GMATPREP is to read the explaination of every question that you did wrong and look at the various types of questions. Prepare some strategy for each question type.

yes .. GMATPrep is the closest to GMAT but it is NOT GMAT ... may be they tweaked the algorithm for GMATPrep to randomly consider some of the questions as experimental while calculating the practice scores....

Is it possible that they have increased the number of experimental questions following the scoretop fiasco ??

perhaps the uncounted/counted thing lowered your quant score as well ... just a guess ... we would never know ...

Interesting analysis, tks marshpa!

I guess if all my questions are counted, maybe the score would be 35-37? still have room to improve.

I give myself maximum two days to think about what happened. Then I will start prepare for retake on Oct 13.

My personality - just do not like to give up.