In October, the College Board released four full practice tests for the new digital SAT. With the digital SAT launching this March for all international students, these new tests are arriving just in time for students to start prepping.
But whether you’re a student who’s eager to start preparing or your theoretical test date is still far off, and you’re simply curious whether the new test is a good fit for you, these four tests may seem like simultaneously too little and too much—too little practice to truly master the new test, and too much material to consider just to get a sense of how (or whether) to take the digital SAT.
Don’t worry! We’ve got you covered.
For the last few weeks, our curriculum team has been hard at work, taking each of these new tests dozens and dozens of times, dissecting and cataloguing every possible module and test question, and pushing College Board’s testing app to its very limit. And even though the College Board has been fairly transparent about the nuts and bolts of the digital SAT (releasing over 200 pages of information last June), there was still a lot for us to uncover. Over the next few blog posts, we’ll dive into the new test’s different sections. To start, here were the three biggest things we learned about how the test works as a whole.
1. How, exactly, the adaptive modules work.
Here’s what we already knew: The new test is split into two sections—first is Reading and Writing (RW); second is Math. Each of these sections is divided into two separately timed “modules” (composed of 25 scored and 2 unscored questions in RW, 20 scored and 2 unscored questions in Math). For each section, there are two different second modules a student might see—a lower-difficulty one and a higher-difficulty one.
Beyond just difficulty level, there were other differences between the possible second modules. Certain concepts were more common on either the higher- or lower-difficulty module. For instance, higher-difficulty second modules for RW contained three times as many questions on punctuation, and they tested concepts like modifiers and parallelism that didn’t show up at all on the lower-difficulty modules; lower-difficulty modules had more than three times as many questions on verbs. None of the higher-difficulty second modules contained questions on poetry. All of this comes with the caveat that just four tests isn’t a large enough sample set for us to say for sure whether these trends will be exactly the same on the actual test.
How well a student does on a section’s first module determines the difficulty level and average point value of questions in the second. In short, if you do really well on the first set of questions, you get harder questions on more challenging topics that are worth more points.
In RW, qualifying for the higher-difficulty second module requires correctly answering approximately two-thirds of the questions in the section’s first module. If you receive the lower-difficulty second module, the highest score you can earn will likely fall somewhere in the upper 500s. Receiving the higher-difficulty second module will peg your minimum score around the 470–500 range. So, if you get the higher-difficulty second module but then answer zero questions correctly, you should score at least in that range.
In Math, qualifying for the higher-difficulty second module requires correctly answering between one-half and two-thirds of the first-module questions. If you receive the lower-difficulty second module, the highest score you can earn varies somewhat—but in most cases is in the 540–550 range. Receiving the higher-difficulty second module will increase your minimum score to approximately the 440–470 range.
You might be asking, why so approximate? Why so vague? We would love to list clear, specific cut-offs for how many questions you need to answer to get a certain score or receive a certain module, but, well…
2. Even within the same module, some questions are worth more than others.
For instance, missing just Q1 might lose you 10 points; missing just Q2 might lose you 20 points. Some missed questions might not subtract at all from your score, unless they’re paired with certain other missed questions. Missing more easy questions than hard questions will also lead to a very different cutoff for reaching the higher-difficulty second module. The technical name for this process is Item Response Theory (IRT), and it’s essential to how the new test works.
Why do it this way? Well, the only way for the new test to be a full hour shorter than the paper-based test, yet still measure student performance on the same scale, is to have an adaptive difficulty—essentially, the test “learns” how proficient a student is and adapts its content and scoring accordingly. By using IRT, the test is a lot “smarter” in its decisions about how to adapt its content: your score and which second module you see doesn’t just depend on how many questions you miss, but which ones you miss.
3. Yes, the test feels easier in many ways, but the scoring isn’t.
Throughout 2022, the College Board has invited certain students to take pilot tests of the new SAT—from these tests, two anecdotal trends emerged:
- Right after taking the test, many students said the new test felt a lot easier and more manageable than the old SAT (something which the College Board has made a big deal about in its initial PR push).
- Then, when scores were released, a greater number of students than usual reported being surprised by lower-than-expected scores.
Both trends align with what we already expected from this new test: There’s less obvious friction in the test-taking experience—the passages are shorter, multi-question passages are gone, the calculator is extremely user-friendly, and there’s more average time per question—so the test should definitely “feel” easier to many students. But, because it’s a shorter test, a single mistake can be more costly than would be on the paper-based test. On the paper test, 20 points is the most that any single question is “worth” (and that’s on the absolute, most punitive scoring scales we’ve seen). On the digital test, we found that missing a single question might decrease your score 30, 50, or even 60 points—depending on which question you miss. If you think of an SAT score in these terms, as the result of a conversion rate from number of misses to amount of points lost, it sure seems like the test is getting more harsh. Here’s a different way to look at it: instead, think of the final SAT score as a conversion from the percentage of questions right. When looked at that way, the scoring scales for the paper and digital SATs aren’t all that different.
The bottom line is that, at least so far, the early results of College Board’s pilot tests show that students are scoring similarly on the digital and paper tests. And the score outcomes and module cutoffs we’ve discussed here are still just scraping the surface of how the actual test’s scoring algorithm will ultimately work. We’ll know more in early 2023, once the College Board releases the full results of its validity study.
In our next post, we’ll dig into the most dramatic changes to the Reading and Writing section. In the meantime, if you have specific questions, reach out to our experts here. We’re happy to help in any way we can.
ArborBridge is the global leader in innovative, digital, one-on-one tutoring. With nearly a decade of experience teaching students online, ArborBridge supports students of all kinds: home schoolers, AP students, test preppers, and more. Our tutors specialize in creating personalized plans and in providing compassionate support for students and families.