ACT Science common mistakes
ACT Science · Understand the traps and how to avoid them.
Reading the passage's introductory text in full before looking at the questions, spending several minutes on background prose that most questions never test.
Candidates transfer habits from reading-heavy exams and assume comprehension of the whole passage is required; ACT Science passages are data-interpretation scaffolds, and the intro often contains only framing, so front-loading reading burns time that the question set does not reward.
What to do instead
Skim the intro in under 20 seconds for the stated purpose and variable names, then go straight to the questions and return to the passage only to locate the specific figure, table, or sentence each question points to.
Treating the graph's axes as interchangeable, e.g., reading a value off the vertical axis when the question asks for the independent variable, or misreading a descending axis.
Under time pressure candidates pattern-match to the shape of the curve and skip the 3-second check of axis labels, units, and direction; descending or non-zero-origin axes are deliberately used to separate careful readers from fast ones.
What to do instead
Before answering any graph question, verbally (silently) state the x-axis label with units and the y-axis label with units, and note whether either axis starts at zero or runs downward.
Answering a two-experiment comparison question using data from only one experiment, or assuming a trend from Study 1 carries into Study 2 without checking.
The first table encountered anchors attention, and candidates generalize from it; ACT Science frequently designs the second study to change one condition (temperature, concentration, species) so the trend differs or reverses.
What to do instead
For any question mentioning two studies or two figures, physically place a finger on each source and confirm the answer is consistent with both before selecting; if the studies disagree, the answer must reflect the disagreement.
Confusing correlation with causation when a question asks what the data 'suggest' or 'support', choosing an answer that claims one variable causes another.
Everyday reasoning treats co-varying quantities as cause and effect, and causal-sounding answer choices are written to be attractive; observational data in the passage cannot establish mechanism.
What to do instead
Default to 'is associated with' or 'varies with' phrasing; reject any choice containing cause/effect verbs (causes, produces, due to, because of) unless the passage describes a controlled manipulation that isolates the variable.
Misreading the question stem's quantifier, answering 'which trial had the greatest value' when the stem asked for the smallest, or 'all EXCEPT' as 'all'.
Stems are long and the operative word is buried mid-sentence; the brain completes the familiar pattern ('greatest') rather than reading the actual word, and NOT/EXCEPT questions invert the selection rule.
What to do instead
Circle or underline the operative word in every stem (greatest, least, EXCEPT, NOT, increase, decrease) before touching the answer choices, and for EXCEPT questions mark each choice T/F against the data and pick the lone outlier.
Interpolating or extrapolating beyond the plotted data range and selecting a value the graph does not actually support.
The trend line looks smooth and the answer choices offer plausible numbers just past the last data point; candidates assume linear continuation without evidence, but the passage only supports values within the measured range.
What to do instead
Check that the requested x-value lies between the first and last plotted points; if it falls outside, eliminate numeric choices and look for a 'cannot be determined' or qualitative answer.
Skipping the figure legends, footnotes, and table captions, then missing that units differ (mg vs. g, minutes vs. hours) or that a symbol denotes a different condition.
Legends are set in small type and feel like decoration; candidates go straight to the numbers, but ACT Science routinely hides the key to a question (a unit conversion or a symbol definition) in a one-line note under the figure.
What to do instead
Read every caption and footnote under each figure and table before using its data, and write any unit conversion factor in the margin so it is applied consistently across all choices.
On conflicting-viewpoints passages, evaluating which scientist is 'right' rather than which statement each scientist would endorse.
Candidates bring a truth-seeking stance and personal prior knowledge, so they pick the view that matches what they already believe instead of tracking the logical commitments of each hypothetical scientist.
What to do instead
For each scientist, write a one-line summary of their core claim and their evidence; then for each answer choice, ask 'would Scientist 1 say this?' and check it against that one-line summary only, ignoring real-world correctness.
Running out of time because low-yield questions are fought to a standstill, leaving the final passage (often the easiest, most data-light one) unattempted.
Candidates treat all questions as equally weighted and refuse to abandon a hard item; ACT Science is paced at roughly 5 minutes per passage, so a single 3-minute stall costs an entire later passage.
What to do instead
Cap any single question at about 45 seconds: mark a guess, flag it, and move on; do a final sweep to fill every blank, since there is no wrong-answer penalty.
Changing a correct answer because an answer choice 'sounds more scientific' or matches outside knowledge rather than the passage's data.
Distractors are deliberately written with technical vocabulary and real-world plausibility, and candidates who are unsure default to the most sophisticated-sounding option instead of the one the figure supports.
What to do instead
Justify every selection by pointing to a specific number, row, or sentence in the passage; if you cannot point to it, the choice is wrong regardless of how plausible it sounds, and outside knowledge should never override the given data.