Everything the adaptive question bank can teach and test in Precise Expression, from foundations through advanced practice. Work through it in order, or start practicing and let the questions find your level.
Treat the two as a trade rather than a rule: before deleting, ask what a reader could no longer rule out. If the answer is nothing, the words were padding; if something, you have swapped one defect for another.
Diagnosing which kind you have is what makes the repair fast — a word with two senses needs a different word, while a phrase that could attach in two places needs restructuring. Applying the wrong remedy leaves the sentence broken.
Reading is reconstruction, so anything you leave underdetermined gets filled in from the reader's assumptions rather than yours. The cost is invisible to the writer, who already knows which reading was intended.
Rather than asking whether a sentence is clear — which you cannot judge about your own writing — try to build a second reading that the words permit. Succeeding is the signal; failing honestly is the finish line.
Stating more than you can support is a claim about your evidence, not about the subject, so it misleads even when every detail happens to be true. Readers calibrate trust on the weakest such claim they catch.
Raising formality tends to reach for longer, more abstract vocabulary, which usually widens rather than narrows the meaning. Assuming the elevated word is the exact one is how formal prose becomes harder to pin down.
An underdetermined sentence lets the writer be read charitably later, whichever way things turn out, so the imprecision is doing work for them and against the reader. Noticing who benefits is the diagnostic.
B. Denotation, connotation, and near-synonyms
Two words can be substitutable in a definition and still deliver opposite verdicts, because the attitude a word carries travels with it into every sentence. Checking the definition alone will not catch the swap.
The verdict rides along whether or not you intended it, so choosing among near-equivalents is an evaluative act. Where you want no verdict, the neutral option usually has to be built from a plainer description.
What a thesaurus omits is the constraint: how strong the word is, how formal, whether it implies someone chose to act. Swapping on the list alone routinely changes one of those without the writer noticing.
Usage fixes which words travel together, independently of grammar, so a legal pairing can still read as wrong. A reader who notices the oddity will look for a reason for it, and find one you did not intend.
Softening changes how heavily the thing lands, which can protect a reader or protect the writer. The test is whether the milder term still lets someone act on what actually happened.
Every accurate label sits somewhere on a scale of severity, so selecting the harshest one argues while appearing merely to describe. It is the mirror of softening, and just as invisible in a fact-check.
A broad word plus qualifiers asks the reader to intersect several constraints; one word whose range already matches hands them the meaning directly. It is usually shorter, but the gain is in the precision.
Reporting verbs encode how strongly the evidence bears on the claim, so swapping one for a livelier alternative silently restates your confidence. Match the verb to what the study or source can actually carry.
C. Specificity and the abstraction ladder
The same thing can be named at many heights, and each step upward discards detail while sounding more authoritative. Writers drift up the ladder under pressure, which is why abstraction correlates with evasion.
Each of the three preferences reduces the number of things the reader could be picturing. They compound, so a sentence can be fixed by moving down any one of them rather than all three.
These words occupy a noun slot without filling it, so the reader supplies the real noun from context or gives up. They are a useful search target precisely because they are so easy to find mechanically.
A category makes the reader do the work of imagining a member; naming the member does it for them and grounds the sentence in something checkable. It also tends to be the shorter option.
An assertion asks for agreement, while the underlying detail lets a reader reach the judgment themselves — and a reader who reached it holds it more firmly. It is also falsifiable, which the assertion is not.
Detail is calibrated by the decision the reader has to make, so the same fact needs different resolution in a summary and in a procedure. Including everything you know is a failure to choose.
D. Wordiness and redundancy
Opening on a placeholder subject pushes the real subject and the real verb later, so the reader holds an empty frame until the content arrives. The repair usually surfaces a stronger verb that was already there.
A doubled phrase performs emphasis without adding a distinction, and readers who notice the repetition look for the difference between the two halves. Finding none, they discount the emphasis.
The modifier restates something the noun already entails, so it adds length while implying a contrast that does not exist. The tell is that the opposite version is impossible rather than merely unusual.
These words signal that the writer wanted more force and did not find the word that carries it, so they mark the gap rather than filling it. Readers discount them automatically, which weakens the sentence.
Each phrase leaves a relation open until the next one resolves it, so stacking them forces the reader to hold several incomplete attachments at once. Converting one into a possessive or a verb usually clears the queue.
These are fixed phrases that each carry a single logical relation, so the long form buys nothing the one-word form lacks. They accumulate in drafts because they are easy to type while thinking.
An opener that announces the sentence rather than making it postpones the claim and implicitly ranks it, which is a judgment the reader would rather form themselves. Deleting it almost never loses content.
The construction wraps a clause in a noun phrase so it can occupy a slot, which almost always survives being rewritten as a simple subordinate clause. It is a reliable marker of a sentence built in the wrong order.
E. Conciseness as a revision procedure
Cutting first and restoring second inverts the usual bias, because defending a word you have already removed is harder than defending one still on the page. What comes back is what was doing work.
Two adjacent sentences about one subject repeat the subject and then need a connective to state a relation that subordination could carry structurally. Merging removes both at once.
A relative clause carries its own subject and verb, and where those add nothing the same relation survives as a phrase. The check is whether the reader can still recover who did what.
Placing an idea in the main clause tells the reader it is the point, so the arrangement argues before the words do. Two sentences with identical content can therefore make different claims.
Take each qualifier and ask what you would say if challenged on it; a defensible answer means it is carrying real uncertainty, and no answer means it is protecting the writer. There is nothing in between.
Without tracking what meaning survived, shortening becomes its own goal and correct qualifications get deleted for being long. Comparing both numbers is what keeps the pass honest.
F. Nominalization and hidden verbs
Converting the action into a noun leaves the sentence needing some other verb to hold it, and the substitute is almost always weaker than the verb that was displaced. The energy of the sentence goes with it.
Look for the action hiding inside a noun and promote it back to the main verb; the weak verb, the preposition and often an article all become unnecessary in the same move. The sentence shortens as a side effect.
These verbs carry almost no meaning of their own, so their presence signals that the real action has been parked in a following noun. Searching for them is a fast way to find the pattern at scale.
Once the action is a noun, the sentence no longer has a slot that demands an actor, so the actor disappears without anyone deciding to remove them. That side effect is why the pattern thrives where responsibility is uncomfortable.
Stacked nouns drop the prepositions and verbs that would have stated how each relates to the next, leaving the reader to guess the structure. Each additional noun multiplies the possible readings.
When the process itself is what you are discussing — comparing it, classifying it, tracking it over time — it needs to be a noun to occupy a subject slot. The pattern is a defect only when the action is the point.
G. Voice and agency
Putting the actor in the subject slot means the reader learns who is responsible in the same instant they learn what happened, with no inference required. Order does the work that would otherwise need an extra clause.
Unlike the active, this structure is complete without naming anyone, which makes it the natural home for sentences where the actor is inconvenient. The grammar gives no signal that a party is missing.
The construction concedes an event while leaving the responsible party unnamed, so it reads as an admission and functions as a deflection. Readers recognize it, which is why it now carries its own connotation.
Try appending the actor; if the sentence improves, the actor was omitted rather than genuinely unknown, and if it reads as redundant the omission was fine. It converts a stylistic hunch into a test.
A procedure is a sequence of assignments, so a step that does not say who performs it leaves the reader unsure whether to act. This is where the structure does the most damage per sentence.
H. When the passive is the precise choice
Where the actor is genuinely unrecoverable, naming one would assert something you do not know, so the structure is the accurate choice. The question is always whether the omission reflects the evidence or the writer's convenience.
Holding one topic in the subject slot across several sentences is what makes a paragraph read as a single line of thought, and sometimes only this structure allows it. Cohesion can outrank the default preference.
Readers expect the newest, heaviest information last, and rearranging the sentence to put it there is a legitimate reason to reverse the usual order. The choice is about information flow, not about avoiding an actor.
In method reporting the identity of the person performing a step is deliberately beside the point, because the claim is that anyone following it would get the same result. The convention encodes that.
A form of be followed by an adjective or a state is not this construction at all, so a search that matches on the auxiliary alone returns false hits. The test is whether an actor could be appended.
Rewriting every instance converts correct sentences — unknown actors, topic chains, end-focus, method reporting — into worse ones. A rule applied without the diagnostic does more damage than the pattern it targets.
I. Tense, aspect, and time precision
Choosing a tense asserts when something held, so an unconsidered choice commits you to a timeline you may not mean. It is checkable in a way most style decisions are not.
One presents the event as closed and located in time; the other presents it as bearing on the present. Where the relevance to now is the point, the two are not interchangeable.
Reporting what someone said shifts the embedded tense in regular ways, and departing from the pattern changes whether the claim is attributed to them or asserted by you. The grammar carries the attribution.
Readers build a timeline from tense, so drifting between them mid-passage leaves them unable to order the events. The damage compounds because each shift invalidates the frame built from the previous one.
One leaves the possibility open and the other states that it did not happen, so the form alone tells the reader whether the case is live. Mixing them asserts something about the world you may not intend.
Marking an action as ongoing rather than finished is how readers infer whether something is still in progress, which they then act on. The distinction often matters more than the tense itself.
J. Modality, hedging, and commitment
These words sit at different points on a scale, so substituting one for another moves your commitment without changing anything else in the sentence. Readers are sensitive to the position even when writers are not.
The same modal can express what someone is required to do or how confident you are that something is so, and only context separates them. Ambiguity between the two readings is a real failure mode in rules.
Where uncertainty exists, reporting it is the accurate move, and stripping the hedge overstates what you know. Confidence is not the same thing as precision, and often the opposite.
A hedge added to make the sentence hard to falsify protects the writer while giving the reader nothing to work with. Ask whether it reports a real range or merely creates room to retreat.
Layering several markers of uncertainty produces no readable level of confidence, because the reader cannot combine them into an estimate. One calibrated hedge conveys more than three vague ones.
Words that assert the reader already agrees substitute for the evidence that would earn agreement, and a reader who does not agree is now also being told they should. They frequently produce resistance.
Match the force of the wording to the strength of what supports it, in both directions — overclaiming and underclaiming are the same error. Doing this consistently is most of the skill.
K. Quantifiers and vague measurement
These words map to ranges that differ between readers and between contexts, so two people can agree on the sentence and disagree on the fact. A number, or a stated range, removes the gap.
A proportional change says nothing about magnitude without the starting point, so the same ratio can describe a large effect or a negligible one. Reporting both is what makes the claim checkable.
A comparative needs a second term to mean anything, and omitting it leaves the reader to supply a flattering one. The sentence then feels like a claim while asserting nothing verifiable.
How many digits you report tells the reader how finely you measured, so carrying extra ones claims a precision your instrument or sample never had. Rounding is part of the claim, not presentation.
A single figure discards the uncertainty that produced it, which readers then cannot recover; an interval carries both the estimate and how much to trust it. The second half is usually what they need.
Readers convert these into rates, and the rate they pick varies by person and by domain, so the same word can mean weekly to one reader and yearly to another. State the rate where it matters.
These are not interchangeable: some indicate a value near the figure in either direction, while others cap one side and permit anything below it. Choosing loosely changes what you have promised.
L. Comparison precision
The second term is where the information lives, so omitting it produces a sentence that sounds like a claim and cannot be checked. Advertising relies on this, which is why readers are primed to notice it.
A comparison only works if both sides are the same kind of thing, and mismatching them produces a sentence with no coherent reading even though the grammar is fine. Name both sides explicitly to catch it.
Every comparative is relative to a reference set, and leaving it implicit lets each reader choose their own. Stating the set is often the difference between an impression and a fact.
The two forms make different claims about how many things were in play, so the slip quietly changes the scope of what you asserted. Count the items before choosing the form.
Some adjectives describe a state that does not admit degrees, so modifying them for intensity signals that a different word was wanted. The construction reads as imprecision even to readers who cannot name the rule.
When the two sides differ in grammatical shape the reader pairs the wrong elements before noticing anything is wrong, and the resulting misreading looks like your claim. Matching the forms prevents it structurally.
M. Modifier placement and scope
Attachment follows proximity rather than intent, so a modifier sitting beside the wrong candidate will be read against it. The sentence is usually still grammatical, which is why the error survives proofreading.
An opening phrase looks for its subject in the main clause, and if the intended one is absent it binds to whatever is in the subject slot. The result is often absurd, and occasionally merely wrong.
Positioned between two plausible targets, the modifier gives the reader no basis for choosing, so the sentence genuinely carries both readings. Moving it to one side resolves it without rewording.
This word restricts whatever follows it, so each position in the sentence produces a different true statement. Writers place it by rhythm and readers interpret it by position, which is where the mismatch comes from.
An adverb can qualify the whole statement or just the action, and placement is what tells the reader which. The two readings can differ in whether you are describing an event or evaluating it.
One version narrows which things you are talking about and the other adds detail about all of them, so the two describe different sets. Punctuation alone often carries the distinction.
When a modifier precedes a coordinated pair, nothing in the structure says whether it reaches the second item, and no punctuation fixes it. Repeating the modifier or splitting the phrase is the only reliable repair.
N. Pronoun reference
Once two nouns of the same number are available, the pronoun no longer selects, and readers resolve it by proximity or by what makes sense to them. It survives editing because the writer knows which one was meant.
Pointing at a preceding stretch of text leaves its boundaries undefined, so the reader must decide how much is included — and different readers decide differently. Adding a noun after it fixes the scope.
The further a pronoun sits from its noun, the more competing candidates have appeared in between, so the link weakens even without ambiguity. Repeating the noun costs a word and removes the work.
Number and person let a reader discard impossible antecedents without thinking, which is why agreement is a precision device rather than an etiquette rule. Where it fails, resolution becomes guesswork.
Following an enumeration, a pronoun has the whole list, its final item and the surrounding discussion all available as candidates. Naming the intended one is usually shorter than the repair for being misread.
The same word can be a structural placeholder with no referent at all, or a genuine pronoun pointing at something. Readers parse both fluently until a sentence makes the two available at once.
O. Singular they and generic reference
It is the established option for a person whose identity is unspecified, and for a person who uses it, so it solves the generic case without imposing a gender the sentence never established. Major style guides endorse it.
Using one gendered form for everyone leads readers to picture that group, which is a factual error about who is meant rather than only a question of courtesy. Studies of comprehension show the effect directly.
The paired form is accurate about two options and heavy to repeat, and it still frames the field as containing exactly those two. It is a workable fallback rather than a default.
Recasting in the plural removes the need to choose a singular pronoun at all, which is why it is often the cleanest repair. It fails only where the singular is load-bearing to the claim.
For a specific person the correct pronoun is the one they use, which makes it a matter of getting a fact right. Treating it as a style preference is the category error.
Changing approach partway through leads readers to look for a reason, and the most available explanation is that a different group is now meant. Pick one strategy and hold it across the document.
P. First person, stance, and attribution
Marking which claims are your own separates them from what you are reporting, and the impersonal alternative deletes that boundary. Readers then cannot tell your conclusion from the literature's.
This pronoun has several available referents and the sentence rarely settles which, so readers resolve it by expectation. The reading they pick changes who is committed to the claim.
Speaking as an institution rather than a person is a choice about who is answerable, and it conceals whether one individual or a body stands behind the statement. That concealment is sometimes the purpose.
Avoiding first person usually produces a construction with no actor at all, which loses the attribution the first person was carrying. The rewrite costs more precision than the convention was worth.
A reader needs to know whether a statement is yours, a source's, or a summary of disagreement, and nothing in a bare declarative says which. Marking it is a precision requirement, not a citation formality.
Fields differ on this, and the convention encodes how that field handles attribution rather than any judgment about self-effacement. Reading the norm as modesty leads writers to strip attribution they needed.
Q. Punctuation that changes meaning
Adding or removing that mark changes whether the clause selects a subset or describes the whole set, so it alters what the sentence is about. It is the rare punctuation choice that is truth-conditional.
Its value shows only in the cases where the last two items could be read as a single unit, and those cases are unpredictable in advance. Using it consistently costs nothing and removes the class.
Two complete statements joined this way leave the relation between them unstated, so the reader supplies one. Choosing a connective or a stronger mark is what makes the relation part of the sentence.
It signals that two complete statements belong together without naming how, which is right when the relation is obvious and wrong when it is the point. Where the reader might guess wrong, say it.
It sets an expectation that what follows will deliver what the first part announced, so a colon followed by something unrelated breaks a commitment the reader was holding. Anticipation is the mechanism.
Both interrupt, but one raises the interruption's prominence and the other lowers it, so the choice ranks the material rather than merely enclosing it. Readers feel the ranking even when they do not notice the marks.
Joining two words tells the reader they modify jointly rather than in sequence, which can change which noun each attaches to. The two versions frequently describe different people or things.
Enclosing a term signals distance from it, and readers correctly take that as a judgment about the term or whoever uses it. Reaching for them to mark an unfamiliar word delivers a verdict you did not intend.
R. Sentence structure and clarity
Material wedged between the two forces the reader to keep the subject in memory until the verb arrives, and the longer the wait the more likely they lose it. Agreement errors cluster here for the same reason.
Everything before the main clause is held in suspense, so a long approach delays the claim past the point where a reader is still building structure. Moving some of it after the verb resolves the queue earlier.
Coordinated items are matched by form first, so unmatched shapes cause the reader to pair the wrong elements before any meaning is computed. Repairing the form usually repairs the reading automatically.
When the subject cannot perform the verb, the sentence has grammatical shape and no coherent proposition, so the reader silently substitutes a plausible one. It survives because nothing looks wrong at the surface.
Switching grammatical patterns mid-sentence leaves no complete parse, so readers reconstruct an intention rather than reading a claim. It happens most often when a sentence is revised from the middle outward.
Heavy elements placed last let the reader finish the frame before absorbing the complexity, while placing them early forces them to hold structure and content at once. The same words are easier in the other order.
Boolean relations written in prose leave grouping unspecified, and in rules or criteria the different groupings are genuinely different requirements. Lists or explicit restatement are the only reliable fixes.
S. Cohesion: given before new
Readers take the opening of a sentence as a statement of what it is about, so putting unfamiliar material there leaves them without an anchor. The sentence can be perfectly clear and still feel disconnected.
The closing position carries emphasis whether or not you intended it, so whatever lands there is what the reader retains. Sentences are often fixed by moving the right element to the end rather than by rewording.
Opening with what the reader already holds and closing with what is new links each sentence to the last, which is what produces the feeling of flow. The rule operates below conscious notice.
Sentences sharing a subject read as developing one argument, while shifting subjects reads as a list of separate observations. Readers infer the structure of your thinking from this pattern.
These words each assert a specific relation — contrast, consequence, simultaneity, addition — so a mismatched one instructs the reader to expect something the next sentence does not deliver. The misdirection is worse than omitting it.
The opening sentence sets a scope, and readers check every following sentence against it, registering anything outside as a digression. Writing the scope you actually cover is cheaper than defending the drift.
T. Jargon, terms of art, and definition
One has a fixed technical meaning that a plain paraphrase would lose, while the other is shorthand a different word could replace. Which is which depends on who is reading, not on how technical it sounds.
A definition folded into the sentence at first use costs one clause and prevents every later misreading, without sending the reader elsewhere. It works best where the term is central rather than incidental.
Expanding once helps only if the reader still holds it when the short form reappears, which long or skimmed documents cannot assume. Re-expanding at section boundaries is cheaper than being misread.
A metaphor drawn from another domain works while the source is shared, and drifts as it spreads, so the same term ends up meaning different things to different teams. Outside readers get nothing from it.
Technical-sounding vocabulary makes a claim feel measured while the underlying statement stays vague, so the form of precision substitutes for the substance. It is harder to challenge than plain vagueness.
A word is precise for a reader who holds its meaning and opaque for one who does not, so the vocabulary decision belongs to the audience rather than the subject. The same sentence can be right for one and useless for the other.
U. Consistency of terms
Fixing a single name per idea removes an entire class of ambiguity at no cost in length, and it is mechanically checkable. It is the highest-return decision available in a long document.
Swapping in a different word for variety leads the reader to look for the distinction it marks, because in careful writing a change of word usually signals a change of thing. The freshness costs more than it buys.
Over a long document near-synonyms accumulate until nobody can say whether three terms name three concepts or one, and by then the answer is unrecoverable. Catching it early is much cheaper.
Marking a term visually tells the reader which words carry a stipulated meaning rather than an ordinary one, so they know when to look it up. Without the signal they cannot tell that a definition exists.
A glossary serves the reader who arrives mid-document and needs to look something up; an inline definition serves the one reading straight through. Which you need follows from how the document is used.
Every inconsistency in the source multiplies downstream, because a translator must decide whether the variation was meaningful and each decision propagates. Consistency is therefore a cost borne by others.
V. Register and tone
The levels differ in vocabulary, syntax and how much shared context they assume, so moving between them changes more than tone. Treating register as a single formality dial hides the syntactic half.
These features travel together, which is why raising formality tends to bring abstraction and lost actors along with it. Knowing they cluster lets you raise the register without accepting the whole package.
They lower formality while leaving meaning untouched, which is why most current style guidance permits them in all but the most formal registers. Avoiding them does not make a sentence more exact.
Addressing the reader directly makes it unambiguous who is expected to act, which is the decision that matters most in instructions. The impersonal alternative reintroduces the missing-actor problem.
A register that does not fit the situation is read as carrying an attitude — irony, condescension, evasion — because readers assume the mismatch was chosen. The content is then filtered through that reading.
A joke depends on context the reader must already share, so it fails first for the readers with the least of it. Those are usually the readers who most needed the sentence to be plain.
One tells the reader to act and the other reports what someone does, so they serve different documents even when describing the same step. Mixing them within a procedure leaves the reader unsure whether to follow along.
W. Bias-free and person-first precision
A broad umbrella collapses groups that differ in the ways the sentence is about, so naming the actual one is more informative before it is anything else. The courtesy and the accuracy point the same way.
Placing the person ahead of the characteristic keeps the characteristic from standing in for them, which matters because a label used as a noun tends to absorb everything else about the referent.
Some communities prefer the characteristic first, treating it as part of who they are rather than something they have. The accurate choice is the one that community uses, which makes it checkable rather than a matter of taste.
Including a characteristic the claim does not depend on implies to the reader that it is relevant, because writers are assumed to mention only what matters. The implication is false and was never stated.
Report at the resolution your data supports rather than collapsing to a familiar category, since the collapse silently asserts that the members are alike in the relevant way. Aggregation is a claim.
Accepted terms move, so a superseded one signals when the text was written and pulls attention onto the wording. The reader's attention is then on you rather than on the claim.
X. Commonly confused pairs, as classes
Sounding identical means these pass every spelling check, so they must be resolved by asking what job the word does in the sentence. They survive proofreading precisely because nothing flags them.
These pairs split by word class, so the question is which slot the sentence needs rather than which meaning you want. Identifying the slot settles it without recalling any rule about the pair.
The division is whether the verb takes an object, so the sentence itself supplies the answer: look for a thing being acted on. Memorizing the pairs is unnecessary once the test is available.
Each of these pairs turns on whether the noun can be counted individually, which is a property of the noun rather than of the pair. One test resolves the whole family.
These rest on grammatical distinctions — what kind of clause follows, or what case the pronoun is in — rather than on meaning. Applying the grammatical test is faster than recalling usage advice.
Each pair encodes who acts on whom, so choosing the wrong one reverses the participants while leaving the sentence grammatical. Naming the two roles before writing settles it.
These encode which side of the relation is the whole and which the parts, so a swap inverts the structure being described. Identify the whole first and the verb follows.
These distinctions are eroding in general usage, which makes the class worth learning as a class: readers who hold the distinction will see the error, and readers who do not lose the meaning. The pattern outlives any single pair.
Y. Global English and translation-safe phrasing
An idiom's meaning cannot be computed from its words, so a reader who has not met it has no route in — and neither does machine translation. The failure is total rather than partial.
These combinations carry several unrelated senses, and only context selects, so a reader with less context picks wrongly with no signal that they have. A single-word equivalent usually exists.
A reference is precise only for readers who hold the source, and it is opaque rather than merely unfamiliar to those who do not. The writer cannot detect the failure from inside the culture.
Explicit, shallower syntax survives translation and second-language reading, while heavy subordination compounds every ambiguity it contains. Each nested clause multiplies the readings rather than adding one.
Dropping the optional connector is recoverable for readers with strong native intuitions and genuinely ambiguous for others, so the saving falls on the readers least able to absorb it. Restoring it costs one word.
Dates, decimal marks, units and address order all vary by locale, so a bare figure can be read two ways by two readers who each apply their own convention correctly. Spelling out the format removes the class.
Z. Diagnosing imprecision in your own draft
Speech forces you through the sentence in order at a fixed pace, which the eye does not do, so stacked modifiers and lost antecedents surface as places you stumble. The stumble locates the defect.
Deliberately hunting for a second reading is the only dependable way to find structural ambiguity, because a writer reading normally supplies the intended parse automatically. Trying to break the sentence is the method.
Marking each sentence's single claim exposes the sentences carrying two, which is where structure gets overloaded and relations go unstated. Splitting them usually reveals a relation you had not committed to.
Swap the word for a candidate and ask what changed; whatever the swap loses is what the original was carrying. It converts a vague sense that a word is right into something you can state.
Reading only for the main verb and who performs it catches buried actions, missing actors and subjects that cannot perform their verbs in a single sweep. Three defects share one diagnostic.
Marking each claim as checkable or not turns vague quantities and bare comparatives into a concrete list of things to pin down. It separates the editing decision from the research it requires.
Keentune is not affiliated with or endorsed by the organizations whose documentation informs these maps.
All exam, test, and product names and trademarks are the property of their respective owners and are used here for identification and reference only. Keentune is independent study practice — not affiliated with, authorized, or endorsed by any of these organizations.