Understanding the point structure of the GRE: how the test is scored

Davis: Hey everybody, this is GRE Bites. My name is Davis, and I’m an educator with over ten years of experience.

Orion: And I’m Orion, the founder of StellarGRE.

Davis: We’re here to bring you your weekly bite-sized episode on GRE prep and grad school admissions. Check out our top-rated GRE self-study program at stellargre.com. And don’t forget, you can use the code “BITES” for 10% off any membership.

Today, we’ve touched on it briefly in a previous episode, maybe two. But, you know, I think we mentioned once what getting a perfect score looks like in terms of its range. I thought it would be a good idea today to lay out the basic facts of the point structure on the GRE, what each section contains, and how many points each question is worth. Are there any pros and cons to skipping questions or leaving them blank versus guessing? Let’s get into that.

Orion: Great. So, to understand the point structure of the GRE, we first have to understand the structure of the GRE, which looks like this: You begin with the writing section, which is composed of two tasks – the “Analyze the Issue” and the “Analyze the Argument” tasks. That’s the first hour of every administration. After that, there will be 5 20-question sets. Whatever you get first will be random.

So, you’ll either get a quant or a verbal set first, and it will alternate from there. It could either be quant-verbal-quant-verbal-quant or verbal-quant-verbal-quant-verbal. Whichever you get three of, one of those three will be the experimental section, which will be unscored. However, it’s unlikely that you will know which one it is, so you kind of have to give your best on every section. That means, for the verbal, there are two scored 20-question sets, and for the quant, there are two scored 20-question sets. So, your overall verbal and quant scores are based on your performance on those 40 scored questions.

The test is also adaptive. Unlike the GMAT, it’s adaptive by section, which means how you perform on, say, the first scored verbal section will determine the relative difficulty level of the second scored verbal section. There are three different second sections you can encounter: we can call them easy, medium, and hard. To get into the top percentiles of scores, you need to have provoked the hardest second section for both the verbal and the quant.

So, there are 40 points each for verbal and for quant, making it 80 points total in terms of your overall score at the end. To tap into those hardest sections, you generally need to excel on your first verbal and your first quant section, aiming for 20 out of 20.

However, you don’t have to get a perfect score. The thresholds for provoking the hardest second sections vary. For example, to provoke the hardest second quant section, you need to answer at least 15 questions correctly on the first graded quant section. For the verbal, I believe you need to answer 14 questions correctly.

Davis: Great, that’s helpful information. For those first two writing sections, how many points are each of those worth?

Orion: So, before I answer that question, let’s talk about what points are. There are actually three different scores associated with the GRE: raw scores, scaled scores, and percentiles. For example, just a few moments ago, you mentioned that there are 40 points for the verbal section and 40 points for the quantitative section, etc. We can call those the raw scores. Raw scores represent the sheer number of questions you answer correctly, with one point awarded per question. It’s crucial to understand that hard questions aren’t worth more points than easy ones. Similarly, questions that take you longer aren’t worth more than those you answer quickly. Every question within the verbal and quant sections is worth the same one point. Those are your raw scores. However, you are never actually told what your raw scores are.

Davis: Students don’t see their raw score on their score report. You would only ever see your raw scores on your practice tests to have an idea of how that would then translate, right?

Orion: That’s correct. The easiest way to track your own progress as you’re preparing for the test is to track your raw scores, which is the sheer number of questions you’re getting right within the time limit. Obviously, that number should be going up. As your prep progresses, you should be answering more questions correctly within the time limit. Emphasizing “within the time limit” is really, really important. A lot of students don’t practice enough within the time constraint because they just want to feel like they can answer the questions. Unfortunately, they’re preparing for a test they’re not going to take, which often leads to some rude awakenings further down the road.

Davis: Great. So we’ve got the raw score, and then we’ve got the scaled score. We’ve already talked about how there’s like a 260, which is kind of the average. What’s the max score on the GRE? What’s the score threshold that’s considered good enough for grad school?

Orion: Sure. So, scale scores are an arbitrary metric. On the verbal and quant sections, they range from 130 to 170, which is a perfect score, in one-point increments. The mean for the verbal is 150, and the mean for the quant is 152. So, a perfect score would be a combined 340, and the 50th percentile would be right around 300. Now, the issue with scaled scores is that if you were to tell someone who wasn’t familiar with these metrics, “Hey, I got a 168,” they wouldn’t know if that’s good. It’s a meaningless number. It’s just a ruler that ETS created to measure the GRE. So, no one really knows what a scaled score means. That’s why, in combination with your scaled scores, you also receive percentile scores, which provide useful information. Your percentile scores tell you the proportion of people who took the same version of the test that you outscored. For all these scores — raw, scaled, and percentiles — higher is always better.

If you get a 50th percentile score, for example, that means you’re right in the middle: half the people who took this test scored lower than you, and half scored higher. For the 90th percentile, 90% scored lower than you. That’s an impressive score; you’re outperforming 90% of people around the world who are trying to become doctors.

Regarding the writing section, there is no raw score. The scaled score ranges from zero to six in half-point increments. And again, that’s a totally arbitrary metric. Therefore, the percentiles are really the useful information.

Davis: Now, that’s helpful. So, just for people to understand, if I’m remembering this correctly, and you’re reminding me, 260 is the lowest possible combined scaled score, and 340 is the highest possible scaled score. But again, unless you’re very familiar with the point structure of all these, which, if you’re tracking your raw score during practice tests, you might have more familiarity than others, the percentile is really what you’re going to be looking at. That’s the information that will come from an admissions office when they say, “This is the average score of people that get into the program,” if I remember correctly.

Orion: Usually, if admissions programs publish that data, it’s median scaled scores for successful applicants. Now, how to convert raw scores into scaled scores is actually kind of a closely guarded secret. ETS clearly has some kind of equation, where certain, you get, you get more points by provoking harder second sections, for instance. But they don’t publish this equation. Obviously, I’ve been able to more or less reverse engineer what I believe to be something plausible by examining the relationship between raw and scaled scores using the freely available power prep practice tests distributed by ETS. And I have created my own equation based on my best guess of what ETS does within the Stellar system. So you can see your raw scores when you’re taking your practice tests on StellarGRE, and we’ll also convert them into scaled scores. Converting scaled scores into percentiles is also freely available; just Google “GRE percentiles.” And like the first thing that pops up will be a PDF document published by ETS, which will on page two contain a table converting scaled scores into percentile scores. And so that’s not a mystery, but it is a mystery kind of how you go from raw to scaled scores.

Davis: Now, that’s really helpful. One quick note I didn’t realize is how much there is to delve into if you have a technical mind, to truly understand the bones of this. But one thing everyone should know is that there is no penalty for an incorrect answer. We’ve talked about skipping questions, going through on a first pass, then coming back. Ideally, every question should have a checked answer when you finish that section.

Orion: That’s right. We say that there’s no guessing penalty. That’s what it’s called in psychometrics. Unlike the LSAT, where there is a quarter-point penalty for getting a question wrong, on the GRE, there’s no penalty for answering a question incorrectly. This basically means that you should put in an answer for every single question, even if it’s just a wild, random guess. Because, why not?

Davis: That’s great. If you have any more questions about that, please shoot us an email at stellargre@gmail.com. We’d be happy to discuss this question or any others in future episodes.

Thanks, everybody, for tuning in. We’ll be back next week with another bite-sized episode of GRE Bites. If you have a topic you’d like discussed on a future episode, let us know at stellargre@gmail.com. And if you’re ready to take your prep to the next level, check out our top-rated GRE self-study program at stellargre.com. You can use the code “BITES” for 10% off all memberships there. Talk to you soon.

Leave a Reply

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