Scoring Scale for the New GMAT Integrated Reasoning Section – Key Takeaways

For the New GMAT Exam coming out in June 2012, the new GMAT Integrated Reasoning Section (30min, 12 questions) will be scored from 1-8.

Now you ask, “Wait! Doesn’t each question have multiple answers? Like 3 Yes/No questions per actual question?”

That’s right!

In order to get that question right, you have to get ALL of the subquestions correct. One little mistake (due to not fully understanding what was “meant” in the question) is going to COST you! There is NO PARTIAL CREDIT!

Are you scared yet?!

GMAT Score Reporting

For all GMAT exams taken June 5, 2012 and later, the 5 scores reported are as follows:

Key points for Integrated Reasoning

  • Not computer adaptive like the Verbal and Quant section of the GMAT
  • Does not count towards your “800” score; score is separate just like the AWA writing score (which is out of 6)
  • NO Partial Credit; must get all subquestions correct to receive credit for that question
  • Scaled score out of 8; percentile ranking reported (and these %iles up change every month based on data collected from each set of test takers each month)

So, what does this mean?
If there are 12 questions and I get full credit for 10, what is my score?

Well first, it’s complicated.

Your score is going to be scaled.

But, there are also experimental questions. It’s possible that there are 4 experimental questions, in which case 8 count towards your raw score of 8. However, you might see 3 or even just 2 experimental questions, in which case you might have 9 or 10 IR questions that actually count. Your score would then be calculated as # of IR questions correct (for the 8, 9, or 10 that actually count) out of the 8, 9, or 10 questions that actually count. That ratio is then placed on a scale out of 8. The integer value (probably rounded) is then your score for the integrated reasoning section.

A lot of different scenarios can play out for your score. In sample example passes for the GMATPrep Software, it was possible to get a full score of 8 with two or even three questions wrong. It’s likely that those 2 or 3 questions wrong were the experimental questions. One possibility is that 10 of the 12 questions counted toward the IR score – meaning 2 of them were experimental. So then, of the 10 questions you might get 9 out of 10 correct. Well, what does 9 out of 10 convert to when scaled out of 8? Well now that depends on the overall difficulty of those 10 questions you got. It’s possible that getting 1 wrong here still converts to an 8!

On the other hand, it’s also possible that you got 4 experimental questions, so 8 of them actually count. Of those 8, you get one wrong. So you have 7 out of 8 correct. What does that convert to for your raw score? Well, difficulty needs to be factored in. Your overall difficulty might be average and 7 out of 8 might convert to a raw score of 7!

So in the first example getting 1 wrong in the set of 10 questions that count might get you an 8, but getting 1 wrong in this set of 8 questions that count might get you a 7!

1 wrong = raw score 8
1 wrong = raw score 7

It’s unclear whether one wrong gets you an 8 or a 7. It entirely depends on:
1) Which questions you got wrong (real vs experimental)
2) How many questions you received that actually counted (8, 9 or 10?)
3) Overall difficulty of the collection of questions you received

In a few sample runs, we saw the following results in GMAT Prep software.

12 correct => 8 raw score
11 correct => 8 raw score
10 correct => 8 raw score
9 correct => 7/8 raw score (depending on total difficulty level)
8 correct => 7 raw score
7 correct => 5/6 raw score
6 correct => 5 raw score
5 correct => 4 raw score
4 correct => 3 raw score
3 correct => 2 raw score
2 correct => 2 raw score
1 correct => 1 raw score
0 correct => 1 raw score

So it looks like in these sample runs, the questions that we answered incorrectly might mostly have been experimental questions – since 2 or even 3 questions wrong still scored an 8.

But note, the above example is not always the case. You have no idea which ones are the experimental questions. If this set of 12 questions differed in number of questions that actually counted and differed in total difficulty of questions, we could have a different outcome for results. So, this chart is not 100% accurate but since there are so many moving parts, you can use this chart above as a guide.

What else can we draw from this chart? It’s a little bit of gamble whether you have breathing room at the top. Getting 2 or even 3 wrong can get you a full score – just hope that you got the experimental ones wrong, got all really difficult questions, and received more than 8 questions that actually counted toward your score. So it’s true, you don’t need to get everything correct in order to get a full score.

In some cases, as long as you demonstrate competency by getting about 10/12 correct you can get full score (with the caveats mentioned above). Either way, we want to reiterate that this section really shouldn’t be your focus when studying for the GMAT. The focus of your energy should be on the verbal and quant sections of the exam that come AFTER you complete this integrated reasoning section on the actual day of the exam. Verbal + Quant are the real sections that count toward your 800 score, which is what business schools REALLY care about.

The order of the exam is:
1) AWA Essay
2) Integrated Reasoning
3) Then Verbal and Quant.

Don’t let the integrated reasoning hurt your stamina, brain power, and most important confidence going into the important section of the exam!

New Section on the GMAT?

Now the big question is, are the questions hard? I went through a bunch – some of the graphics, data tables are straightforward – maybe a few booby traps but nothing crazy that can’t be handled with good practice. Except for two-part and multi-source reasoning.

When you get to multi-source reasoning which is basically like critical reasoning / reading comprehension and data sufficiency combined (thus, “integrated” reasoning), there’s room for getting confused or not catching some small detail here or there.

So, is the IR section going to help you or be against you? Well, for business schools, it’s definitely going to help them.

The GMAT was originally created to have a mean of 500 and a standard deviation of 100. A score of 600 would be one standard deviation above the mean. However, over the years with the best GMAT prep available and students using smart strategies and thinking processes like those offered by GMATPill, the average score has trended up and the number of students scoring 700+ is much more common these days. And if you’re applying to top schools where nearly everybody has a 700+ score, well, how else can business schools screen?

The application and background of the candidate is very important. But there are plenty of students who make it to the waitlist – and it’s really a tough spot to be in. Candidates are in agony waiting to find out if they get into their dream school. Meanwhile, there is not much they can do. Sometimes getting an improved GMAT score can get them off the waitlist – but this is not guaranteed. The addition of the IR score can help a school make some screens at this point. The IR section uses more real world business questions -and may become increasingly important in admission decisions. But for the time being, the traditional GMAT score out of 800 is going to be most important while the IR score is going to be supplemental and will come in handy in helping business schools screen through hundreds of candidates in agony on the waiting list.

Percentile Ranking

In addition to the IR score reported out of 8, the percentile ranking will be sent to schools. At first, the percentile will be based on only a small sample size of sample test scores done privately before June 5, 2012. Every month thereafter, all test takers’ data will be inputted into the cumulative score data and percentile rankings will be recalculated every month. So while a 7 out of 8 on the exam might get an 83%ile (for exaxmple) in June, the same score in July may turn out to be 93% if everyone does well – OR it may fall to a 76% if everyone has a tough time.

These percentiles will help business schools in their admission selections.

As of June 25, 2012, the most percentile rankings are as follows:

Full Score 8 = 94%
7 = 85%
6 = 70%
5 = 54%
4 = 46%
3 = 26%
2 = 17%
1 = 0%

Source: GMAC

These percentiles will be updated monthly for some time as more data points are collected. Generally, if you are shooting to be score 700+ on the GMAT Exam (which is around 90th percentile for the combined Verbal + Quant score), then you should aim to have a 7 or 8 on the IR section. This means of the questions you got that actually count (remember, we remove the experimental questions), you need to nearly all of them correct. There may be some breathing room to get 1 and maybe 2 questions wrong (out of the questions you got that *actually* count).

Your IR percentile ranking along with score will be accessible to business schools. So don’t skip the preparation, read more about GMATPill’s IR Pill.

Getting Practice

Here at GMATPill, we have you covered. The IR Pill (released!) brings forth thinking processes and strategies that have made the existing pills a giant success – SC Pill, CR Pill, RC Pill, DS Pill and PS Pill.

The IR Pill has launched and is preparing students for the increasingly important IR section of the exam. There are sample practice questions and explanations that we created that you can look at to help you get a feel. You can also try the new Integrated Reasoning questions on the GMATPill Practice Pill Platform.

  • Multi-source Reasoning
  • Two-Part Analysis
  • Table Analysis
  • Graphics Interpretation

Leave a Reply

Your email address will not be published. Required fields are marked *

/images/logo.png" />