Why the easy Craft and Structure questions on the Digital SAT are the ones students mark wrong

The Digital SAT Reading and Writing section is split across four broad content domains, and Craft and Structure is the one that decides whether a student settles at 600 or pushes into the 700s. Of the roughly 54 operative items a test-taker sees in a single sitting, between 13 and 16 of them fall inside this domain, which makes it the single largest contributor to the Reading and Writing scaled score after Information and Ideas. Yet in my experience tutoring students at the SAT preparation programme, Craft and Structure is also the domain where capable readers lose points they did not need to lose. The questions read like comprehension, they feel like vocabulary, and the wrong answers feel reasonable on a first pass. That is exactly why a targeted preparation strategy matters: the surface skill (reading a passage) is not the same as the tested skill (interpreting why the author chose a word, a transition, or a paired source). This article walks through the three Craft and Structure question families, the exact error patterns I see in classroom diagnostics, and the preparation sequence that turns a plateau into a measurable gain inside the adaptive modules.

What the Craft and Structure domain actually tests on the Digital SAT

Craft and Structure is one of the four content domains reported on the Digital SAT Reading and Writing score report. It sits alongside Information and Ideas, Expression of Ideas, and Standard English Conventions. The College Board groups questions here when the test is asking a student to read the architecture of a text, not its surface meaning. In practical terms, the questions fall into three operational families: Words in Context, Text Structure and Purpose, and Cross-Text Connections. Each family probes a different layer of the same underlying competency, which is the ability to explain how a writer's choices shape a reader's understanding.

Words in Context items ask the student to decide which version of a word best fits the precise meaning the author is trying to carry. The stem shows the word inside the sentence with a blank, and the four choices are often close in dictionary definition but very different in connotation or register. The tested skill is not vocabulary breadth in the abstract; it is the ability to read the surrounding clauses and decide which meaning is being invoked. A student who studies word lists to prepare for these items is studying the wrong signal. The signal is in the syntax and the surrounding logic of the sentence.

Text Structure and Purpose items ask the student to explain how a sentence, a paragraph, or a passage is built and why the author made those choices. Typical stems include 'Which choice completes the sentence so that it accomplishes the author's goal?' and 'Which choice best describes the function of the highlighted sentence in the passage as a whole?' The student must read the highlighted unit, then read the surrounding context, then describe the relationship. Cross-Text Connections items appear in the paired-passage format. Two short sources are presented together, and the student is asked how a claim in one source relates to a claim in the other, or which choice best characterises the relationship between the two authors' positions.

Each Reading and Writing module contains a fixed, balanced quota of these three families. If a student has not seen any Cross-Text items by the second half of Module 2, the test is signalling something about the adaptive path, and a tutor reading the score report can use that to diagnose pacing or attention. For the student, the practical implication is straightforward: preparation should drill all three families, not just the one that feels easiest. I have watched strong readers plateau at 640 for months because they practised only Words in Context and treated the structure and cross-text items as something they would 'pick up by reading more books'. The Digital SAT does not reward that strategy.

Words in Context: how to read the sentence before the stem

The most common error pattern in Craft and Structure is the Words in Context trap in which the student picks a synonym of the highlighted word that does not actually fit the logic of the sentence. Consider a sentence such as 'The committee's decision, though defensible on procedural grounds, did not address the underlying concern raised by the faculty.' A student who treats defensible as a vocabulary item will scan for a synonym and may pick reasonable, which is superficially related. But the sentence contrasts the procedural soundness of the decision with its failure to address the real concern; the correct answer is the one that preserves that contrast, often a word like justifiable or tenable, depending on the choices. The point is that the surrounding clauses are doing the work.

The preparation move that works in classroom drills is to train the student to ignore the choices for the first ten seconds. Read the sentence with the highlighted word. Read the sentence with a blank. Then articulate in your own words what the sentence is actually saying, including the relationship between the clauses. Only then look at the four options. If an option does not preserve the relationship you articulated, eliminate it regardless of how close it looks in a thesaurus. This single reordering of attention rescues several points per module for most students.

There is a second trap inside Words in Context that catches even strong readers: the use of an opposite-meaning word as a distractor when the sentence contains a contrast signal such as 'though', 'although', 'however', or 'yet'. The student reads the contrast, picks the antonym, and marks it correct. But the contrast may not be the one the question is asking about. The sentence can contain two relationships, and only one of them is the operative one for the highlighted word. The right preparation move here is to underline every conjunction and conjunctive adverb in the sentence and to ask, for each one, whether the highlighted word sits on the same side of the contrast as the test is asking about. This is a slow technique at first; after about two weeks of daily practice, it becomes fast and automatic.

A third pattern worth naming is the 'register mismatch' distractor. The four choices are sometimes drawn from different registers: a literary word, a colloquial word, a technical word, and a neutral word. Only one matches the register of the surrounding passage. A student who has been reading academic journal articles for class will sometimes pick the technical choice even when the passage is written in a neutral journalistic register. The diagnostic question is: does the surrounding passage sound like a textbook, a magazine, a speech, or a set of instructions? The answer determines the register the highlighted word must match, and that filter eliminates two of the four options before the student even reads the choices carefully.

Text Structure and Purpose: the question family students under-prepare

Text Structure and Purpose questions ask the student to do two things at once: read the highlighted unit, and read the unit that surrounds it, in order to describe a relationship. The relationship can be additive ('the next sentence provides an example of the claim just made'), contrastive ('the next sentence complicates the claim just made'), causal ('the highlighted sentence explains why the previous claim is true'), or sequential ('the highlighted sentence marks a transition from setup to evidence'). The four choices usually present four different relationships, only one of which is the actual relationship in the passage.

The error pattern I see most often is that the student identifies a true statement about the content of the highlighted unit and marks it correct, without checking whether that statement describes the function or whether it merely summarises. A typical wrong choice on a Text Structure item looks like this: 'The highlighted sentence notes that early experiments produced inconsistent results.' That is a true statement. But the question is asking what the sentence does in the passage. The correct answer might be 'The highlighted sentence introduces evidence that the next paragraph will attempt to explain.' The student who marks the first answer is not wrong about the content; they are wrong about the question being asked. This is why I tell students in the SAT preparation programme to read the stem twice and to circle the operative verb: 'best describes the function', 'best completes the sentence so that it accomplishes the author's goal', 'best states the relationship between the first paragraph and the second'.

A practical preparation sequence for Text Structure and Purpose runs as follows. First, take a single passage and write, in plain English, one sentence per paragraph describing what that paragraph is doing in the passage. Not what it is about, what it is doing. Is it setting up a claim? Introducing a counter-argument? Presenting an example? Drawing a conclusion? Second, for every transition word in the passage (however, therefore, for example, by contrast, in practice), write the relationship the transition marks. Third, when answering a Text Structure item, the student should be able to point to the surrounding sentences that justify the relationship they have chosen. If they cannot, the answer is a guess, and a guess inside Craft and Structure is a 25 percent proposition.

There is also a structural sub-family that deserves specific mention: questions that ask the student to choose a transition or a connector that best fits between two sentences. The four choices are usually all grammatically possible; only one preserves the logical relationship the test is targeting. The diagnostic move is identical to the Words in Context move. Translate the relationship in plain English ('the second sentence gives an example of the first', 'the second sentence shows a result of what the first describes', 'the second sentence concedes a point before pushing back'). The choice that matches that plain-English relationship is the correct one. This reframe is a strong preparation move because it removes the student's dependence on memorising transition word lists, which is a low-yield strategy for a test that deliberately varies its distractor phrasings.

Cross-Text Connections: how the paired-passage format actually works

Craft and Structure contains the only paired-passage format in the Reading and Writing section, and it is the format students underestimate the most. Two short sources are presented side by side, typically between 40 and 90 words each, and the student is asked a question that requires reading both. The questions come in two operational shapes: relationship questions ('which choice best describes the relationship between the two passages') and synthesis questions ('based on the passages, which choice best states a claim that both authors would reject').

Relationship questions are the easier of the two shapes, and a strong reader can usually identify the right answer within 30 seconds once they have the diagnostic frame. The four choices are usually labelled by relationship: agrees, disagrees, extends, or qualifies. The student must decide, for each author's central claim, whether the other author's claim supports it, contradicts it, builds on it, or limits it. The error pattern I see is that the student identifies a true claim in passage B and marks the choice that says passage B supports passage A, when in fact passage B contradicts passage A on a different point. The fix is to identify the central claim of each passage first, before reading the choices, and to test each choice against those two central claims.

Synthesis questions are harder because they require the student to hold both passages in mind and to invert the operation: not 'what does each author say' but 'what would each author reject'. A typical stem reads: 'Based on the passages, both authors would most likely agree that…' and the choices are sometimes true statements, sometimes statements only one author would agree with, and sometimes statements both would reject. The student must mentally run each choice through both authors' positions. The preparation move is to take a paired passage and to write, for each author, three claims they would agree with and three they would reject. Then the synthesis items become fast: a choice that one author would reject is wrong, regardless of whether the other author would agree with it.

There is also a pacing consideration specific to Cross-Text Connections. Because the student has to read two passages and answer at least one question about them, the time budget is tighter per item than in Information and Ideas. The SAT preparation programme's pacing recommendation is to spend no more than 90 seconds on a paired-passage question. If a student has spent 90 seconds and cannot narrow the choices to two, the right tactical move is to mark the most defensible remaining option and move on. A guess inside a single Craft and Structure item costs a student roughly 0.2 percent of the Reading and Writing scaled score, and a missed question later in the module that the student would have answered correctly costs far more.

How Craft and Structure interacts with the adaptive module routing

The Digital SAT is a multistage adaptive test, and the way Craft and Structure is weighted changes between Module 1 and Module 2 in a way that matters for preparation. Module 1 contains a mixed-difficulty set of items across the four content domains, and the student's performance on Module 1 determines whether Module 2 is the easier version or the harder version. The easier Module 2 routes a student to a scaled score band that tops out below the threshold needed for the most selective admissions cycles. The harder Module 2 unlocks the upper band of the Reading and Writing scaled score.

Craft and Structure appears in roughly the same proportion in both modules, but the difficulty distribution inside the family shifts. In the easier Module 2, Words in Context items tend to have choices that are clearly different in register, and Text Structure items tend to have relationships that are marked by obvious transition words. In the harder Module 2, the relationships are implicit, the registers are closer, and the distractors are designed to lure a student who has not done the diagnostic work described above. A student who has been practising on easier items will see the harder module and feel a sudden drop, even if their underlying skill has not changed. The fix is to prepare on harder items from the start.

The score report also provides subscore data on Craft and Structure separately, which means a student can see whether their losses are concentrated in Words in Context, Text Structure, or Cross-Text Connections. For most students I tutor, the subscore pattern looks like a single weakness rather than an even spread. A student who scores 14/16 on Words in Context and 8/14 on Text Structure and Purpose is not a student with a general Craft and Structure problem; they are a student with a specific Text Structure problem, and their preparation should reflect that. The SAT preparation programme's diagnostic workflow is built around this kind of decomposition: identify the family, identify the error pattern inside the family, drill the pattern, and re-measure.

Common pitfalls and how to avoid them

Several predictable error patterns show up across the Craft and Structure domain, and a short tactical list helps anchor the preparation plan. The following pitfalls appear in almost every diagnostic I run, and the counter-moves are tested in classroom drills.

• The synonym trap in Words in Context. The student picks a dictionary synonym of the highlighted word without checking the surrounding clauses. Counter-move: read the sentence with a blank, articulate the relationship in plain English, then evaluate each choice against that relationship.
• The 'true but not functional' choice in Text Structure. The student identifies a true content statement and marks it correct, when the question is asking about function. Counter-move: circle the operative verb in the stem; if the stem asks for a function, eliminate every choice that describes content.
• The contrast signal misread. The student reads a contrast conjunction and assumes the highlighted word must carry the opposite meaning. Counter-move: identify the side of the contrast the highlighted word sits on, and verify the choice preserves that side.
• The register mismatch. The student picks a choice whose register does not match the surrounding passage. Counter-move: identify the register of the passage (academic, journalistic, technical, literary) and eliminate choices that do not match before reading them for meaning.
• The Cross-Text central-claim conflation. The student reads passage B and assumes it supports passage A because the surface topic is the same. Counter-move: write the central claim of each passage before reading the choices, and test each choice against both claims.

For most candidates, the highest-yield preparation move is the diagnostic decomposition. Identify the family inside Craft and Structure that is producing the losses, identify the specific error pattern inside that family, and drill the counter-move for that pattern with timed practice. A general 'do more reading' plan will not move the subscore on its own; a targeted plan will.

Preparation sequencing: a six-week plan for Craft and Structure

The preparation sequence I assign to students working through the SAT preparation programme's Reading and Writing module runs six weeks and is anchored to a specific weekly output. Week 1 is diagnostic: the student takes a full-length Reading and Writing section under timed conditions, and we decompose the Craft and Structure losses by family and by error pattern. Week 2 drills the highest-loss family exclusively, with 20 to 30 untimed items per day and a written annotation for every wrong answer. Week 3 introduces timed practice on the same family, at a target of 75 seconds per item. Week 4 brings the other two families back into the rotation, and Week 5 layers in mixed-domain timed practice to simulate the actual module balance. Week 6 is consolidation: full-length Reading and Writing sections with a subscore review after each one.

Within the six-week plan, two tactical routines matter more than the rest. The first is the 'stem circle' routine: every time the student reads a stem, they circle the operative verb (best describes, best completes, best states the relationship). This single habit eliminates the 'true but not functional' error pattern almost entirely within two weeks. The second is the 'central claim first' routine for Cross-Text items: before reading the choices, the student writes the central claim of each passage in one sentence. This forces the diagnostic work to happen before the choices, which is where it has to happen for a 700+ band student.

For students aiming at the 700 to 780 band on Reading and Writing, a useful benchmark is to target 12 to 14 correct out of 15 in a randomly selected Craft and Structure item set drawn from the harder module difficulty. That is roughly 80 to 93 percent accuracy on the family, and it is the operating range that allows the rest of the Reading and Writing score to carry the student into the upper band. Below 80 percent accuracy on Craft and Structure, the scaled score band compresses; above 93 percent, the student is leaving points on the table somewhere else, usually in Expression of Ideas or in pacing on the paired-passage format.

Craft and Structure in relation to the other Reading and Writing domains

It helps to position Craft and Structure against the other three domains so the preparation plan can be balanced. Information and Ideas tests the student's ability to locate, interpret, and draw inferences from claims and evidence. Expression of Ideas tests the student's ability to revise and edit a text for effectiveness, including organisation, cohesion, and rhetorical synthesis. Standard English Conventions tests the student's command of sentence boundaries, punctuation, agreement, and the conventions of standard English usage. Each domain is scored on a 1 to 15 subscore band, and the four subscores are summed to produce the 200 to 800 Reading and Writing scaled score.

Of the four domains, Craft and Structure has the strongest overlap with Information and Ideas, because both require the student to read a passage carefully and to interpret the author's choices. The difference is granularity: Information and Ideas questions ask what a passage says, while Craft and Structure questions ask how and why the passage says it. A student who is strong on Information and Ideas but weak on Craft and Structure usually has a strong general reading habit but a weak analytical habit; the preparation plan should focus on slowing down and naming relationships. A student who is strong on Craft and Structure but weak on Information and Ideas usually has the analytical habit but loses points on inference and evidence-evaluation; the plan there is different.

The table above is a working reference rather than an exhaustive taxonomy, but it gives a student and their tutor a clean way to allocate preparation time. For most candidates aiming at a Reading and Writing band in the 700s, a roughly equal distribution across the four domains is a sensible default, with the distribution tilted toward whichever subscore is lowest on the most recent score report.

Putting it all together: a tutor's reading list and drill sequence

The reading list I recommend for Craft and Structure is short and specific. Two short opinion pieces per week, drawn from serious outlets, paired with two short academic abstracts per week, drawn from open-access journals. The point is to expose the student to a mix of registers and to a mix of explicit and implicit transitions. For drills, the highest-yield resource is a bank of Craft and Structure items sorted by family, with the harder-module items flagged. The student should drill 20 items per day per family for two weeks, then taper to 10 per day as the full-length sections come in.

For the Diagnostic phase, a tutor's most useful tool is a subscore chart. The chart should track the student's accuracy on each of the three Craft and Structure families across each of the last five practice sections, with notes on the dominant error pattern for each loss. A pattern that holds across three consecutive sections is a real weakness, not noise. A pattern that fluctuates is a preparation or attention problem, and the fix there is pacing and review discipline rather than content drill.

For the Consolidation phase, the right habit is the post-section review. After every full-length Reading and Writing section, the student should mark every Craft and Structure item they got wrong, write the family, write the error pattern, and write the counter-move they should have used. The pile of these notes becomes the student's personal error catalogue, and the catalogue is the single most useful document a student can carry into the final two weeks of preparation. The SAT preparation programme's Reading and Writing module is built around exactly this cycle: diagnose, drill, measure, review, drill, measure, and so on, until the subscores settle into the band the student is targeting.

Finally, on exam day, the tactical advice for Craft and Structure is consistent with the broader Digital SAT pacing plan. Read the stem twice. Circle the operative verb. Identify the family. Apply the family-specific counter-move. If 90 seconds have passed and the choice is not narrowed to two, mark the most defensible remaining option and move on. A student who has trained these habits for six weeks will execute them without thinking, and the Craft and Structure domain will become the largest single contributor to a Reading and Writing scaled score in the 700s.

For students targeting the upper Reading and Writing band, the next preparation step is to run a subscore diagnostic on a recent score report and to identify the single Craft and Structure family producing the largest share of losses. SAT Courses' Digital SAT Reading and Writing Module 2 hard-route programme analyses each student's Words in Context, Text Structure and Purpose, and Cross-Text Connections error patterns against the rubric and turns a 700+ target into a concrete six-week preparation plan.

Frequently asked questions