NAEP common mistakes
NAEP · Understand the traps and how to avoid them.
When a reading passage asks for the main idea, the candidate copies a vivid supporting detail or the first sentence of the passage as the answer instead of stating what the whole passage is mostly about.
Under time pressure, test takers latch onto a single memorable line and stop evaluating the passage as a whole; a striking detail feels like a summary because it is easy to recall and quote.
What to do instead
Before looking at options, write a one-sentence paraphrase of the passage in your own words. Then check whether each option covers the whole passage or only one paragraph; eliminate any option that is true but narrow. Confirm the surviving option can be supported by evidence from at least two different parts of the passage.
In reading questions that require citing textual evidence, the candidate picks an answer and then chooses a quotation that sounds related but does not actually support the claim, or picks evidence first and reverse-engineers a claim to fit it.
Evidence questions are often presented as two linked items; candidates answer the first quickly and treat the second as an independent comprehension check rather than a logical test of the first answer.
What to do instead
Treat the pair as one argument. State the claim in your own words, then test each quotation by asking 'if this sentence were deleted, would my claim still be supported?' Choose the quotation that directly proves the claim, and if none does, revisit the claim rather than forcing a match.
On multi-step mathematics items, the candidate computes a correct intermediate value and selects it as the final answer, or drops units and reports a number that answers a different question than the one asked.
Intermediate results appear among the answer choices by design, and working memory load causes the candidate to stop as soon as a familiar-looking number appears.
What to do instead
Before computing, underline what the question asks for and its units. After computing, reread the question and label your result with units, then check whether the label matches the request. If a choice matches an intermediate step, treat that as a warning sign and finish the remaining step.
The candidate applies a memorized procedure to a problem whose structure has changed, for example using a proportional method on a relationship that is not proportional, or assuming a figure's appearance gives its measurements.
Practice builds pattern recognition around surface features such as keywords and diagram shapes; when the underlying relationship changes but the wording stays similar, the familiar procedure is triggered automatically.
What to do instead
Before choosing a method, restate the given relationships in symbols or a table and identify what type of relationship holds (linear, proportional, additive, exponential). Never take lengths or angles from a diagram unless they are marked; redraw or annotate the figure with only the given information.
In science items, the candidate explains a phenomenon by naming a concept ('because of density', 'because of natural selection') without linking the concept to the specific mechanism or evidence in the scenario.
Vocabulary-heavy instruction rewards naming terms, and graders and item writers reward causal chains; candidates substitute a label for an explanation because it feels complete.
What to do instead
Use a because-therefore structure: state the observation, the mechanism that produces it, and the result. Ask whether your answer would still make sense to someone who had never heard the term; if not, spell out the mechanism. When data are given, cite the specific pattern or value that supports the mechanism.
When asked to design or critique an investigation, the candidate proposes a procedure that changes several variables at once, omits a control condition, or fails to state what would count as evidence for or against the hypothesis.
Candidates focus on getting a plausible result rather than on making the test logically decisive; school labs often supply the design, so students rarely practise constructing one.
What to do instead
For every design, explicitly name the independent variable, dependent variable, and at least one controlled variable, and state the comparison group. Add a prediction: 'if the hypothesis is right, we would observe X; if not, Y.' Check that the design could produce a result that would refute the hypothesis.
In writing tasks, the candidate answers the prompt with a general opinion essay, ignoring the specified audience, purpose, or source material, and never revises beyond fixing spelling.
Timed writing encourages a rehearsed five-paragraph template; candidates spend their planning and drafting time producing text and leave no time for rereading against the task requirements.
What to do instead
Spend the first few minutes decoding the prompt into a checklist: audience, purpose, required sources, and required length. Draft with that checklist visible, then reserve the final portion of time to reread once for task fit (does every paragraph serve the stated purpose?) and once for sentence-level errors, reading aloud in your head.
The candidate describes or evaluates a work of art or a performance using only personal taste ('I liked the colours') rather than analysing how specific formal elements, techniques, or context create an effect.
Everyday responses to art are framed as preference, and candidates have had little practice separating description, analysis, interpretation, and judgement as distinct moves.
What to do instead
Use a four-step frame: describe what is literally there, analyse how the elements are arranged (composition, contrast, rhythm, medium), interpret what the arrangement suggests, then judge with criteria you state. Anchor every interpretive claim to a specific observable feature.
In civics items, the candidate confuses the roles and powers of institutions, for example attributing to the president a power held by Congress, or treating a state action as a federal one, and answers from general impressions of how government 'should' work.
Civic knowledge is often absorbed from news coverage that blurs institutional boundaries; candidates reason from the headline-level impression rather than from the constitutional allocation of authority.
What to do instead
Build a one-page reference grid of institutions (Congress, president, courts, states, citizens) against powers (legislate, enforce, adjudicate, amend). For each item, name the actor and the power before evaluating options, and check the level of government involved. Distinguish what the Constitution permits from what is politically common.
In geography and history items, the candidate treats maps, timelines, and documents as illustrations to skim rather than as sources to interrogate, missing scale, direction, periodisation, or the author's perspective.
Candidates read sources for content words that match the question and skip the metadata (legend, date, author, scale) that determines what the source can actually show.
What to do instead
Apply a source protocol before answering: for maps, read title, legend, scale, and orientation; for timelines, locate the event's date relative to the reference point; for documents, identify author, date, audience, and purpose. Then ask what the source can and cannot establish, and eliminate options that overreach the evidence.