SC1 Interpretation of Data

Lesson

Interpretation of Data is the biggest thing on the ACT Science section — ACT reports it as 38–50% of the scored questions. It is also the most mechanical. Almost every item is answered by putting a finger on a specific cell, row or point and reading what is there. You are not being asked what you know about photosynthesis. You are being asked what the table says.

Read the figure before you read a word of the text

The framing paragraph exists to tell you what the variables are. Once you know that, the table does the rest. Before question 1, spend fifteen seconds on each table and answer four questions about it:

• What is in the left column? That is usually the thing the researchers changed.
• What is in every other column? That is what they measured.
• What are the units? mL/h is not mL. °C is not K. mmHg is not atm. Half of all misses are unit misses.
• Which direction does each variable go? Down the column, is it rising, falling, or turning around? Mark the turn with your pencil — you write in the booklet, so use it.

Passages carry far more information than any question needs. A three-table passage may have thirty numbers and six questions. Do not study all thirty. Orient, then let the questions send you back.

The single-value lookup

The most common item type is: find the row, find the column, read the cell. The whole difficulty is bookkeeping. Track the row with the pencil tip in your left hand and the column heading with your eye, and you will not slide one line off. Sliding one line off is the single most common wrong answer, which is exactly why the adjacent value is nearly always one of the four choices.

Describing a trend precisely

"Increases" is often not enough. ACT distinguishes shapes, and so must you:

• Increases steadily — roughly equal steps.
• Increases, then levels off — the last two or three values are nearly identical.
• Increases at a decreasing rate — still rising, but each step is smaller than the last. Subtract consecutive pairs and look at the differences, not the values.
• Increases, then decreases — there is a maximum in the middle. Circle it.
• Decreases to a minimum, then increases — common in ocean and atmosphere data.

The test of a trend claim is arithmetic. If a choice says a variable levels off, the last two rows have to actually be close. If it says the rate of increase is falling, the consecutive differences have to actually shrink. Check the differences; do not eyeball the values.

Interpolation and extrapolation are not the same move

Interpolation is estimating between two measured points. If 400 gives 4.0 and 800 gives 5.1, then 600 gives something between 4.0 and 5.1 — and that is a safe, keyable answer. Extrapolation is guessing past the end of the data, and the data does not license it in the same way. If a curve has been flattening, continuing it in a straight line is wrong.

So when a question asks about a value outside the measured range, the credited answer is usually a bracket ("greater than the last measured value") rather than a specific number. When it asks about a value inside the range, the credited answer is usually a specific number between the two neighbours.

Two figures that share a variable

When a passage gives you two tables, find the column heading they have in common — temperature, depth, concentration, time. That shared variable is the bridge, and nearly every hard question in the set walks across it. The routine is: enter Table 1 with what you are given, come out with the shared variable, then enter Table 2 with that.

Watch for the passage that quietly repeats one condition in both tables. If Table 1 was run at 22 °C and Table 2 has a 22 °C row, those two cells must agree — and the question that notices this is usually the hardest one in the set.

Translating between a table and a graph

Ask which variable is on which axis, then convert three points, not all of them. Take the first row, the last row and the row where the direction turns. If a described graph rises then flattens, the table must rise then flatten. If the described curve and the table disagree at any one of your three checkpoints, that description is out.

Approximation is the default arithmetic

No calculator, one minute per question. So round first and compute second. 336 seconds for 3,000 km is "about 3,000 over about 340," which is a bit under 9 km/s — and the choices will be 4.9, 8.9, 11.2 and 15.0, so "a bit under 9" is already enough. Ask how precise the answer has to be before you decide how precisely to work, and let the spread of the choices tell you.

Useful habits with no calculator: halve and double instead of dividing; turn percentages into fractions; compare by ratio rather than by subtraction when the numbers are large; and when two choices are on opposite sides of the truth, you only need the sign of the difference, not its size.