A free preview, generated by MasteryLoopHQ.
A new study of the SAT's reading section drew attention online this week. Here's what it actually compared and found. The Johns Hopkins Institute for Education Policy's report was released on October 29, 2026, according to the sponsoring college, "for Thomas Aquinas College," which sponsored it. Its title is The SAT Reading Section: A Comparative Analysis of the 1980, 1987, and 2026 Practice Tests. The researchers looked at the verbal section of one 1980 form, one 1987 form (both from the National Archives), and one current College Board paper practice test, called Bundle 6, which the report labels "SAT 2026." That is three forms in all. The report calls itself "a descriptive content analysis, not a statistical test," so no students took anything. Its central question is "how the College Board's conception of college-ready reading has changed." The study also states a limit up front: a content analysis "cannot settle" whether the test got easier or harder.
According to College Board, the digital SAT has two sections. Reading and Writing is 64 minutes and 54 questions, split into two 32-minute modules. Math is 70 minutes and 44 questions. Testing time totals 2 hours 14 minutes, plus a 10-minute break between sections. The test is adaptive: "The first module of each section contains a broad mix of easy, medium, and hard questions. Based on how you perform on the first module, the second module of questions will either be more difficult or less difficult." Students take it in the Bluebook app. Each question follows "short reading passages (or passage pairs) followed by a single multiple-choice question." Passages run 25 to 150 words and come from literature, history and social studies, the humanities, and science. Questions fall into four content domains, including Craft and Structure, which covers vocabulary "in context," and Standard English Conventions. The study used a paper practice form, not the adaptive test. On paper the second module is fixed, and there are 33 questions per module against 27 online. The report calls the paper form "a reasonable but imperfect proxy" for the adaptive test.
One count in the study stands out: "In both the 1980 and 1987 tests, 70.6% of questions (60 of 85) focused on vocabulary, compared to only 15% (10 of 66) in the 2026 test." On the old forms, 45 of those 60 were analogy and antonym items. The other 15 gave some context and asked students to fill in one or two blanks. All 10 on the 2026 practice form "provide sentence context and require filling in only one blank." One piece of context the percentages leave out: 30 of the 2026 form's 66 items are grammar and writing items, which the old verbal forms studied didn't have, so the two percentages divide by different totals. The report separates two things. Analogies and antonyms disappearing is a format change. A "reduction in the demand for independent word knowledge" is a different claim, and the study says its own claim concerns the second. College Board's history is that this has been happening for a while. In 2005 the "new" SAT "replaced analogy items with short passage-based reading comprehension items." The 2016 redesign brought "the elimination of the testing of obscure vocabulary ('SAT words')." College Board's account of Words in Context is that those questions focus on "high-utility academic, or tier two, words," and its example of words left out is "lassitude, supercilious, and adumbrate." There is another view. The study says the old design "would naturally favor students with more extensive reading backgrounds." Some read that as a strength, because the old test rewarded wide reading. Others read it as a fairness problem. Both readings are held.
The study's Table 3 puts the change in numbers. The 1980 form had 6 passages averaging 304 words, about 1,820 words in all, with 60 minutes. The 1987 form had 6 passages averaging 340 words, about 2,040 words in all, with 60 minutes. The 2026 practice form had 17 passages averaging 97 words, about 1,650 words in all, with 64 minutes. The old forms put several questions on each passage, and the new form has one question per passage. In the report's words, "the 2026 test taker has roughly 20-30% more time per word of passage text, and the newer form never asks a student to sustain attention across a 300+ word passage with multiple questions." College Board takes a different position on what shorter passages mean, though its comparison is with the paper-based SAT Suite that preceded the digital test, not with the 1980s forms. Anticipating the question of whether shorter passages reduce rigor or mean stamina isn't valued, its framework answers, "The answer to both these questions is a firm "no."" It writes that the digital Reading and Writing section "is no less rigorous than the paper-based SAT Suite," and that SAT scores, paper or digital, "are not intended to directly measure students' reading stamina." A College Board research brief from July 2026 also looked at passage length. With a sample of nearly 200,000 students, it "find[s] that reading passage length does not meaningfully affect student performance on SAT reading and writing questions." The Johns Hopkins report responds in a footnote that the brief "addresses a different question from ours." That brief looks at scores, not content, and compares the recent paper SAT, not the 1980s forms. The report says its own finding "concerns only how much text the test asks students to read."
This is the finding that complicates an "easier" story. Using official Lexile measures from MetaMetrics, the study reports 1128L for the 1980 form, 1337L for 1987, and 1428L for the 2026 practice form. Flesch-Kincaid grade levels run 11.1, 15.2 and 16.4. The study's conclusion is that the 2026 passages are "at least as complex as, and by the Lexile measure more complex than, the older ones." It keeps several caveats, and so should we. Each era is a single form. The gap between 1980 and 1987 (209L) is more than double the gap between 1987 and 2026 (91L). Both measures lean on sentence length and are less stable for very short texts, and the report says the 16.4 grade level is "best read as an artifact." Its safest reading is "the 2026 passages measure above the 1980 form and near the 1987 form." Its plain-English summary: "the newer passages are shorter but denser: a student reads far fewer words, yet the sentences are harder and the vocabulary rarer once reading begins."
The study also classifies passages by type. Argumentative or persuasive passages went from 5 of 12 on the old forms to 1 of 17 on the 2026 form. Expository passages, which explain, went from 5 of 12 to 14 of 17. Fiction was one to two passages per form. Table 5 counts topics: the 2026 form covers 11 knowledge domains, against 8 for 1980 and 7 for 1987. On background knowledge, the report says the newer passages "provide more of the necessary specialized context within the text itself." The older ones often assumed "broader historical, cultural, or scientific frameworks," which "could favor those already familiar with certain topics or perspectives." College Board says plainly that "Topic-specific prior knowledge is not assessed." Here are two readings of the same fact. Some see it as a fairness gain: a student who hasn't been exposed to a topic can still work the passage. Others see it as a loss: the test no longer rewards the knowledge that strong readers build over years. The Johns Hopkins report itself names both. It calls the equity advantages of reduced reliance on assumed background knowledge "real," while saying the new demands (chart reading and writing) "warrant continued scrutiny of their fairness and construct coverage."
What the Johns Hopkins report concludes. Its comparison was one 1980 form, one 1987 form, and one current paper practice test (Bundle 6). Four of its five findings, taken together, "point toward the newer form asking less of the test taker; the fifth—text complexity—points the other way." It adds writing and chart tasks, where "the newer test plausibly asks more." It also says its findings "point to the test becoming less demanding along the dimensions we examined." And it says, "We therefore do not claim that the SAT as a whole has become easier." Its summary is that the test has "redistributed its demands — away from accumulated vocabulary and assumed background knowledge and toward in-passage writing and data interpretation." What College Board says. It describes the digital Reading and Writing section as "no less rigorous" than the paper-based SAT Suite, and says the goal of the redesign was to keep "the value, rigor, and predictive power of the paper-based SAT Suite." College Board has not, as far as we found, responded to this report. A College Board pilot study at 12 four-year colleges, with 1,990 students who were academically strong, found digital SAT scores "predict college performance as well as paper and pencil SAT scores." That compares the digital test with the recent paper SAT, not with the 1980s forms. What critics say. The college that sponsored this study worries that the test no longer rewards wide vocabulary, sustained reading and background knowledge. The college's vice president and dean emeritus, quoted in its news release, said, "What it now measures is different from what it measured in the past." Can you compare scores across eras? Not directly. In April 1995 College Board recentered the score scale so that the average for a 1990 reference group sat at about 500. The Johns Hopkins report advises colleges to "be cautious when comparing scores across eras." The report recommends a study in which "the same students respond to items from both old and new forms," and the sponsoring college says it would welcome one. Whether the changes are progress or loss depends, as the report says, "on what one takes the SAT to be measuring." If you think it should measure accumulated knowledge and stamina, the changes look like a loss. If you think it should predict college grades fairly for students from different backgrounds, they look like a reasonable redesign. Both views are held by serious people.
The report's advice to families is: "Sustained reading across science, history, and the arts remains the most durable preparation." College Board says something similar about its own test: "knowledge building in the subject areas lays the foundation for success on the tests and, more importantly, in test takers' postsecondary educational pursuits." At home, that can look like this. First, read whole things: a long article, a chapter, a speech, not only excerpts. Second, mix subjects: one history piece, one science piece, one essay that argues a point, which is the type the test now uses least. Third, meet words in context, then own them: when a word stops you, work it out from the sentence first (that's what the test asks), then check it, then use it in your own sentence. Fourth, explain it back: after reading, say in your own words what the author claimed and why, and if you can't, reread. This builds the understanding that vocabulary and background knowledge grow from. This isn't a strategy for beating the SAT. It's how readers get stronger, which helps on any version of the test.
A Johns Hopkins study compared one 1980 SAT form, one 1987 form, and one current paper practice test. The current form has far fewer vocabulary questions (10 versus 60, all now in sentence context) and much shorter passages (about 97 words versus 304 to 340). But its passages measure as complex as the 1987 form's or more, it covers more subjects, and it adds writing and chart tasks. College Board says the digital section is "no less rigorous" than the paper SAT it replaced. The report doesn't claim the whole test got easier; it says the test's demands were "redistributed." Whatever you conclude, the report's advice to families is wide, sustained reading across subjects.
In your own words: (1) What did the Johns Hopkins study actually compare, and what can a content analysis not tell us? (2) Name two ways the 2026 practice form asks less than the 1980s forms, and one way it asks as much or more. (3) Why might one person call the background-knowledge change "fairer" and another call it a "loss"? Which view do you lean toward, and why?
Reading a good explanation feels like understanding it. Usually it isn't the same thing — and you don't find out which one you've got until someone asks you to explain it back.