When the SAT went digital in 2024, the reading section split into shorter passages with one question each. The conventional wisdom was that vocabulary mattered less — passages were shorter, context did more lifting. That turned out to be half-true.
The new format killed sentence-completion-style “obscure word” questions. What it kept — and arguably amplified — is the academic vocabulary that signals comprehension across both Reading and Writing modules. Words like ambivalent, equivocal, empirical, pragmatic, exacerbate. These appear regularly across digital SAT reading and writing passages and they’re the words students most often gloss over without realizing they’ve missed the author’s stance.
I built this deck around 50 academic words that recur across released-form digital-SAT material and most reliably index whether a student will read a passage’s tone correctly.
Why these 50 and not the 1000-word lists you’ll find online
Most SAT vocab lists are derived from old paper-test analysis. A lot of those words (pulchritude, quotidian, salient) tested vocab-as-vocab. The digital test rarely does that. What it does — constantly — is hinge a question on whether you correctly read an author’s qualified stance.
That makes vocab clusters more valuable than alphabetical lists. Ambivalent / equivocal / nuanced / measured / circumspect form one cluster: all of them describe an author hedging. Categorical / unequivocal / emphatic / decisive / unambiguous form the opposite cluster: an author committing. If you can place every academic word into one of these stance-clusters, your reading score moves up by a band.
The cluster structure
Hedge words (10 cards). ambivalent, equivocal, nuanced, measured, circumspect, qualified, tentative, provisional, contingent, hedging.
Commitment words (10 cards). categorical, unequivocal, emphatic, decisive, unambiguous, unwavering, resolute, adamant, intransigent, dogmatic.
Skepticism words (8 cards). dubious, incredulous, skeptical, dismissive, derisive, disparaging, supercilious, condescending.
Endorsement words (8 cards). laudatory, effusive, ardent, fervent, sanguine, enthusiastic, sympathetic, partial.
Empirical-language words (8 cards). empirical, anecdotal, corroborative, refutative, conclusive, indicative, suggestive, definitive.
Outcome-modulating words (6 cards). exacerbate, mitigate, ameliorate, attenuate, exacerbating, intensifying.
That’s 50. They cluster into six conceptual buckets that map cleanly onto the stance-detection question type that dominates the digital reading section.
What the cards look like — cloze, not Q&A
Vocab cards are usually built as Q&A: front says ambivalent, back says having mixed feelings. That format works for the first 100 words and then degrades — students learn to recognize the word in isolation but can’t read it in a sentence, which is exactly what the digital SAT tests.
The deck uses cloze deletion instead. Each note embeds the word in a sentence that signals its stance bucket:
An author who is "{{c1::ambivalent::mixed feelings}}" about a topic
neither fully endorses nor rejects it; this is a {{c2::hedge}} stance,
not a {{c3::commitment}} stance.That’s three cards from one note. Card 1 hides ambivalent — student must produce the word from the definition + bucket label. Card 2 and 3 reinforce the cluster taxonomy (hedge vs. commitment) which is the actual skill the digital SAT tests.
The commitment cluster card uses the overlapping pattern:
Words that signal author commitment — full endorsement with no hedging —
include {{c1::categorical}}, {{c2::unequivocal}}, and {{c3::adamant}}.Three cards from one note. Each card hides one commitment word while showing the other two as anchor points. Cluster cohesion is reinforced as a side-effect of the cluster review.
For words with subtle distinctions, the ::hint syntax disambiguates partial recalls:
Someone who is "{{c1::dismissive::short, contemptuous}}" rejects an
idea without serious engagement; "{{c2::derisive::mocking}}" rejection
adds active mockery on top.The hint lets the student check whether their guess landed on the correct shade of meaning before flipping the card.
How spaced repetition changes the timeline
Vocabulary is the textbook spaced-repetition use case. The reason students hate vocab study is that the cramming timeline (memorize → forget → re-cram → forget) doesn’t match how memory actually works.
FSRS-5 — the algorithm Avorio uses — schedules each card based on how well you specifically recall it. The hedge words you grew up with show up every 30+ days. The empirical-language cluster you’ve never used in a sentence will show up every 2-3 days until it sticks. By exam day, every word in the deck has been reviewed at the exact interval that produces the longest retention with the least total effort.
Six months before the test date is a comfortable start. Twenty cards in week one, twenty more in week two, the last ten in week three. After that, ten minutes a day for the rest of the run-up. By week 24 the reviews have stretched out to where most days you’ll see 4-6 cards, which takes 60-90 seconds.
Try the first 10 below
The deck embedded below is the hedge cluster — ten words that describe an author qualifying their stance. If you can place all ten correctly into “hedge” rather than “commit” or “skepticism,” you don’t need this deck. If 3+ are uncertain, install Avorio and load the full 50.