Orient to the figure before you read any prose
Every figure carries the same small set of information, and extracting it takes about ten seconds. What is the independent variable - the thing deliberately changed? What is the dependent variable? What are the units on each? What do the title and legend say? Answering those four questions turns a wall of numbers into a readable object.
Do this for each figure before touching the questions. The prose describes how the experiment was run, which matters for a minority of questions, but the figures are where the answers live. And note the units deliberately: a question asking for grams when the table reports milligrams is testing whether you looked.
- Identify the independent variable, the dependent variable, and both sets of units.
- Read the title and legend before any data point.
- Orient to every figure first, then let the questions send you to one.
Tables: find what changes and what is held constant
In a table, the leftmost column is usually the variable that was changed and the rest are what was measured - worth confirming rather than assuming. Read the header row carefully, units included, then scan one column top to bottom to see which way the numbers move.
Take a generic example. Suppose a table lists water temperature against dissolved oxygen: 10 degrees Celsius with 11.3 mg/L, 20 degrees with 9.1 mg/L, 30 degrees with 7.6 mg/L, 40 degrees with 6.5 mg/L. As temperature rises, dissolved oxygen falls, so the relationship is inverse. Asked for the value at 25 degrees, you would interpolate between the 20- and 30-degree rows and answer around 8.4 mg/L; the choices are normally spread widely enough that a careful estimate lands cleanly on one. When a table has several measured columns, check whether they move together or oppositely, since comparison questions target that.
- Confirm which column is the manipulated variable instead of assuming.
- Read the units in the header row before reading any value.
- Interpolate between neighboring rows for values not listed.
Graphs: axes, scales, and the traps in them
Graphs carry the same information as tables plus one hazard: the scale. Check where each axis starts, since an axis that does not begin at zero can make a small change look dramatic. Check whether the spacing is even or logarithmic, because on a logarithmic scale equal distances mean equal multiples. And check that each axis runs in the direction you expect - a variable such as depth may increase downward.
Watch for graphs with two vertical axes, one on each side, each belonging to a different plotted line. Reading a curve against the wrong axis produces a confidently wrong number, and the legend is your only guide to the pairing. For values between plotted points, read along the curve; beyond the plotted range, extend the established trend, but stay cautious - far extrapolation is where wrong choices wait.
- Check the starting value and the spacing on each axis before reading data.
- With two vertical axes, match each curve to its own axis using the legend.
- Read between points along the curve; extend beyond the data cautiously.
Name the trend in one precise sentence
A large share of questions test only whether you can describe a relationship correctly. Train yourself to state it in one sentence: as the independent variable increases, the dependent variable increases, decreases, or stays roughly constant. Said in that form, the matching choice usually becomes obvious.
Not every relationship is monotonic, and the exceptions are where careless readers lose points. A curve may rise to a peak and then fall; the correct description names the turning point rather than only the first half. Separately, distinguish a trend from a single comparison - 'which sample had the highest value' is a lookup, while 'how did the value change as concentration increased' is a trend. Answering one with the other is a common miss.
- Say it as one sentence: as X increases, Y increases, decreases, or stays flat.
- Watch for peaks and minimums; name the turning point, not half the curve.
- Separate lookup questions from trend questions before answering.
Multi-experiment passages: track what changed
Some passages present several related studies. The organizing question is always the same: what did the researchers change between Study 1 and Study 2, and what did they hold constant? That single difference is the reason the second study exists, and many of the questions are built on it.
Note the difference in a few words as soon as you spot it - 'Study 2: same setup, higher temperature'. That note tells you which study a variable was tested in and prevents the most common error here, reading a value from the right table in the wrong study. Design questions often ask why a condition was held constant, and the answer is nearly always about isolating one variable: holding everything else fixed is what lets a difference in results be attributed to the one thing that changed.
- Identify what changed and what stayed constant between studies, and write it down.
- Check which study a question refers to before reading any value.
- Constants and controls exist to isolate one variable - usually the answer to design questions.
Cross-referencing and two-part answer choices
The hardest data questions combine two figures: read a value from the table at a given condition, then carry it into a graph to read a second variable. Do it in two deliberate steps and write the intermediate value down instead of holding it in your head - rushing that hand-off is where an answerable question turns into a guess.
You will also meet answer choices with two parts - a yes-or-no verdict plus a because-clause. Split them: settle the verdict from the data first, which eliminates about half the options, then evaluate only the reasoning attached to the survivors, since choices frequently pair a correct verdict with reasoning that contradicts the figure. A related type gives you a claim and asks whether the data support it; translate it into a concrete prediction - 'the values should rise as concentration rises' - and check the figure against that.
- For two-figure questions, write the intermediate value down before continuing.
- On verdict-plus-reasoning choices, settle the verdict first, then judge the reasoning.
- Turn any claim into a concrete prediction about the figure and test it.
Practice the reading, not the science
Because the section tests figure reading rather than content knowledge, practice with whole passages and their figures, and check your reading against the explanation for every question - including the ones you got right, since a lucky guess and a correct reading look identical on a score sheet. Then review by error type: a misread axis or unit, a trend named too loosely, a value pulled from the wrong study, or bad reasoning accepted behind a correct verdict. Each has a distinct fix, and the same type recurs until it is named. The ACT Science passages in ScoreMint come with the tables and figures the questions depend on and a per-choice explanation for every answer, and misses return through spaced review.
- Check your reasoning on correct answers too, not just on misses.
- Sort errors: axis or unit, loose trend, wrong study, or bad reasoning accepted.
