GMAT Strategy

Scorers with quant 50 are often interested what it takes to get a quant score of 51. In this thread I will post results from my experiments with GMATPrep to clarify the answer.

The experimental result is:
* You can answer wrong only 4 tough questions on GMATPrep and still get a Quant score of 51 (98th percentile).
* Answering two easy questions before question 10 and one tough question after that gave a score of 50 (92nd percentile). That is not always reproducible, depends on how easy the easy questions were. The easier they are, the more they hurt your score.
* Answering 5 questions wrong (3 average + 2 tough) gave a score of 50.

The point is that if you want a quant score of 51:
* you have to answer all easy questions correctly or you may skew the scoring algorithm into giving you too easy questions later and not be able to earn a 51 score. So you have to work on your precision on easy and intermediate questions, under time pressure. Easy questions usually happen in the begining of the test, when most people tend to make most mistakes.
* you have to learn to recognize and NOT fall in the well-known traps that GMAT sets up, usually in DS problems:
- verbal traps: a single modifier word changes the meaning of the quant question
- integer context: when a problem is IMPLYING integer-only solutions vs all real numbers, that resticts the possible solutions severely and often you end up with a single solution in DS problems which is sufficient. Some problems will say "integers" explicitly, in other problems involving number of people, of objects, of tickets, .... the integer solutions are implied because these variables cannot be non-integers or negative. A positive context is implied in geometry problems where the lengths cannot be negative.
- a system of two equations and two unknowns in which the second equation turns out to be a proportional version of the first one, basically having a single equation for two unknowns, which is insufficient
- a single equation for two variables but being able to solve for a combination of variables that is of interest in the problem, which is sufficient
- even power equations usually have more than one solutions and will be insufficient
- even power equations with additional restriction that rejects one of the solutions will be sufficient
etc.
* you have to zoom on the right answer of DS problems using the AD/BCE system, without reading the actual answers. That saves time and will prevent you from getting the right answer but marking the wrong one.
* you have to learn to do high-level problems with plugging in numbers, especially maximization/minimization problems with restrictions. There is no clear cut algebraic approach for those.

All of the above applies to GMATPrep, which we know is a very good estimator of your score on the actual exam. By induction, it should apply to the actual exam too.

Experiment 1: 4 tough questions wrong (14, 30, 33, 34) quant score 51
------------------------------------------------------------------------
Using GMATPrep version 2.1.279 on Windows 7 64 bit

Notes: the toughness of the questions is subjective, I have no access to the test internal algorithm to see what toughness was assigned to each question. That is why I selected questions that I consider tough and answered them wrong, trying to avoid questions in the begining, in a row, because I've already seen that may swing the scoring towards lower difficulty questions later.

I am posting the wrongly answered questions so that everyone can judge the difficulty for himself/herself:

14.
The number 75 can be written as the sum of the squares of 3 different positive integers. What is the sum of these 3 integers?
(A) 17 (B) 16 (C) 15 (C) 14 (E) 13
Correct ans: E

30.
If p is a positive odd integer, what is the remainder when p is divided by 4?
(1) When p is divided by 8, the remainder is 5.
(2) p is the sum of the quares of two positive integers.
Correct ans: D

33.
If r < s < t, is zero halfway between r and s?
(1) s is to the right of zero.
(2) The distance between t and r is the same as the distance between t and -s.
Correct answer: C

34.
A certain city with a population of 132,000 is to be divided into 11 voting districs, and no district is to have a population that is more than 10 percent greater than the population of any other district. What is the minimum possible population that the least populated district could have?
(A) 10,700 (B) 10,800 (C) 10,900 (D) 11,000 (E) 11,100
Correct ans: D

Experiment 2: 1 easy (2) + 3 hard (7, 31, 32) questions wrong quant score 51
---------------------------------------------------------------------------------------------

The test shows that even if one of the 4 wrong answers is on an easy question, you can still get a 51 score.

The wrongly answered questions were:

2.
If the speed of x meters per second is equivalent to the speed of y kilometers per hour, what is y in terms of x? (1 kilometer = 1,000 meters)
(A) 5x/18 (B) 6x/5 (C) 18x/5 (D) 60x (E) 3,600,000x
Correct ans: C

7.
A certain city with a population of 132,000 is to be divided into 11 voting districs, and no district is to have a population that is more than 10 percent greater than the population of any other district. What is the minimum possible population that the least populated district could have?
(A) 10,700 (B) 10,800 (C) 10,900 (D) 11,000 (E) 11,100
Correct ans: D

31.
Is m + z > 0 ?
(1) m - 3z > 0
(2) 4z - m > 0
Correct ans: C

32.
If Bob produces 36 or fewer items in a week, he is paid x dollars per item. If Bob produces more than 36 items in a week, he is paid x dollars per item for the first 36 items and 3/2 times that amount for each additional item. How many items did Bob produce last week?
(1) Last week Bob was paid a total of $480 for the items that he produced that week.
(2) This week Bob produced two items more than last week, and was paid a total of $510 for the items that he produced this week.
Correct ans: E

Experiment 3: 2 intermediate (7, 13) + 3 hard (18, 28, 35) questions wrong; quant score 50
-----------------------------------------------------------------------------------------------------------------

Apparently GMATPrep does not tolerate 5 wrong questions and drops the score from 51 to 50. In any of my other tests, I was not able to get a quant score 51 with 5 or more wrong questions, so 5 wrong questions seems to be a threshold between scores of 51 and 50.

The wrongly answered questions were:

7.
If n and y are positive integers and 450y = n^3, which of the following must be an iteger?
I. y/(3 x 2^2 x 5)
II. y/(3^2 x 2 x 5)
III. y/(3 x 2 x 5^2)
Correct ans: I only

13.
For every integer k from 1 to 10, inclusive, the k-th term of a certain sequence is given by (-1)^(k+1)/2^k. If T is the sum of the first 10 terms in the sequence, then T is
(A) greater than 2
(B) between 1 and 2
(C) between 1/2 and 1
(D) between 1/4 and 1/2
(E) less than 1/4
Correct ans: D

18.
If x is positive, which of the following could be the correct ordering of 1/x, 2x, and x^2?
I. x^2 < 2x < 1/x
II. x^2 < 1/x < 2x
III. 2x < x^2 < 1/x
Correct ans: I and II only

28.
Last month 15 homes were sold in Town X. The average (arithmetic mean) sale price of the homes was $150,000 and the median sale price was $130,000. Which of the following statements must be true?
I. At least one of the homes was sold for more than $165,000.
II. At least one of the homes was sold for more than $130,000 and less than $150,000.
III. At least one of the homes was sold for less than $130,000.
(A) I only (B) II only (C) III only (D) I and II (E) I and III
Correct ans: A

35.
Tanya arranged 4 different letters to be sent to 4 different addresses. For each letter, she prepared an envelope with its correct address. If the 4 letters are to be put into the 4 letters at random, what is the probability that only one letter will be put into the envelope with its correct address?
(A) 1/24 (B) 1/8 (C) 1/4 (D) 1/3 (E) 3/8
Correct ans: D

Experiment 4: two easy (2, 7) + 1 hard (29) questions wrong; quant score 50
-----------------------------------------------------------------------------------------------------------------

This is an example how the scoring algorithm can be swung towards lower scores just from making mistakes on two easy questions in the begining. So be espeically careful not to make mistakes on those.

The wrongly answered questions were:

2.
A certain office supply store stocks two sizes of self-stick notepads, each in 4 colors: blue, green, yellow, or pink. The store packs the notepads in packages that contain either 3 notepads of the same size and the same color or 3 notepads of the same size and of 3 different colors. If the order in which the colors are packed is not considered, how many different packages of the types described above are possible?
(A) 6 (B) 8 (C) 16 (D) 24 (E) 32
Correct ans: C

7.
Last year a certain bond of a face value of $5,000 yielded 8 percent of its face value in interest. If that interest was approximately 6.5 percent of the bond's selling price, approximately what was the bond's selling price?
(A) $4,063 (B) $5,325 (C) $5,351 (D) $6,000 (E) $6,154
Correct ans: E

29.
The number 75 can be written as the sum of the squares of 3 different positive integers. What is the sum of these 3 integers?
(A) 17 (B) 16 (C) 15 (C) 14 (E) 13
Correct ans: E

Interesting post..

Hi,

A request: Can you please hide all the answers using spoilers. Nobody will be benefited if they see the answers before even attempting the question.

Keep up the good work.

tutorphd wrote:Scorers with quant 50 are often interested what it takes to get a quant score of 51. In this thread I will post results from my experiments with GMATPrep to clarify the answer.

The experimental result is:
* You can answer wrong only 4 tough questions on GMATPrep and still get a Quant score of 51 (98th percentile).
* Answering two easy questions before question 10 and one tough question after that gave a score of 50 (92nd percentile). That is not always reproducible, depends on how easy the easy questions were. The easier they are, the more they hurt your score.
* Answering 5 questions wrong (3 average + 2 tough) gave a score of 50.

The point is that if you want a quant score of 51:
* you have to answer all easy questions correctly or you may skew the scoring algorithm into giving you too easy questions later and not be able to earn a 51 score. So you have to work on your precision on easy and intermediate questions, under time pressure. Easy questions usually happen in the begining of the test, when most people tend to make most mistakes.
* you have to learn to recognize and NOT fall in the well-known traps that GMAT sets up, usually in DS problems:
- verbal traps: a single modifier word changes the meaning of the quant question
- integer context: when a problem is IMPLYING integer-only solutions vs all real numbers, that resticts the possible solutions severely and often you end up with a single solution in DS problems which is sufficient. Some problems will say "integers" explicitly, in other problems involving number of people, of objects, of tickets, .... the integer solutions are implied because these variables cannot be non-integers or negative. A positive context is implied in geometry problems where the lengths cannot be negative.
- a system of two equations and two unknowns in which the second equation turns out to be a proportional version of the first one, basically having a single equation for two unknowns, which is insufficient
- a single equation for two variables but being able to solve for a combination of variables that is of interest in the problem, which is sufficient
- even power equations usually have more than one solutions and will be insufficient
- even power equations with additional restriction that rejects one of the solutions will be sufficient
etc.
* you have to zoom on the right answer of DS problems using the AD/BCE system, without reading the actual answers. That saves time and will prevent you from getting the right answer but marking the wrong one.
* you have to learn to do high-level problems with plugging in numbers, especially maximization/minimization problems with restrictions. There is no clear cut algebraic approach for those.

All of the above applies to GMATPrep, which we know is a very good estimator of your score on the actual exam. By induction, it should apply to the actual exam too.

Hmm,

Great Insight. Thanks

Interesting.Have u done similar research for what separates Q50 ppl and others?

Thanks,
Shreeraj

I got Q50 in quant. only two wrong q17 and q25 of gmat prep. why?

tutorphd wrote:Scorers with quant 50 are often interested what it takes to get a quant score of 51. In this thread I will post results from my experiments with GMATPrep to clarify the answer.

The experimental result is:
* You can answer wrong only 4 tough questions on GMATPrep and still get a Quant score of 51 (98th percentile).
* Answering two easy questions before question 10 and one tough question after that gave a score of 50 (92nd percentile). That is not always reproducible, depends on how easy the easy questions were. The easier they are, the more they hurt your score.
* Answering 5 questions wrong (3 average + 2 tough) gave a score of 50.

The point is that if you want a quant score of 51:
* you have to answer all easy questions correctly or you may skew the scoring algorithm into giving you too easy questions later and not be able to earn a 51 score. So you have to work on your precision on easy and intermediate questions, under time pressure. Easy questions usually happen in the begining of the test, when most people tend to make most mistakes.
* you have to learn to recognize and NOT fall in the well-known traps that GMAT sets up, usually in DS problems:
- verbal traps: a single modifier word changes the meaning of the quant question
- integer context: when a problem is IMPLYING integer-only solutions vs all real numbers, that resticts the possible solutions severely and often you end up with a single solution in DS problems which is sufficient. Some problems will say "integers" explicitly, in other problems involving number of people, of objects, of tickets, .... the integer solutions are implied because these variables cannot be non-integers or negative. A positive context is implied in geometry problems where the lengths cannot be negative.
- a system of two equations and two unknowns in which the second equation turns out to be a proportional version of the first one, basically having a single equation for two unknowns, which is insufficient
- a single equation for two variables but being able to solve for a combination of variables that is of interest in the problem, which is sufficient
- even power equations usually have more than one solutions and will be insufficient
- even power equations with additional restriction that rejects one of the solutions will be sufficient
etc.
* you have to zoom on the right answer of DS problems using the AD/BCE system, without reading the actual answers. That saves time and will prevent you from getting the right answer but marking the wrong one.
* you have to learn to do high-level problems with plugging in numbers, especially maximization/minimization problems with restrictions. There is no clear cut algebraic approach for those.

All of the above applies to GMATPrep, which we know is a very good estimator of your score on the actual exam. By induction, it should apply to the actual exam too.