ACT Science — experiments, controls and what a result will not support

The research summaries on the ACT Science section describe experiments and then ask what those experiments actually show — which is almost always less than the confident-sounding options suggest. This material trains the judgment that separates a result from a conclusion.

Ten questions on described experiments: seedlings on two watering regimes, an alloy heated with and without a coating, two brands of battery, two catalysts, algae under changed light, a toy car on two floors, and pairs of trials to be read together. Every description and every number is written out inside its own question, so nothing depends on a figure or a table.

Two questions ask which variable was deliberately changed and which was held the same. The second is harder than it looks, and its options are split on purpose between things the researchers SET and things they MEASURED — a quantity you measure at the end cannot also be a control, even if it happens to come out similar in both groups.

Two ask what a comparison group is for. The alloy question makes the point concretely: coated bars staying solid at 500 degrees proves nothing on its own, because the alloy might have survived that heat uncoated. It is the uncoated bars melting that turns a result into evidence.

Two ask what a result supports, and their wrong options are the four standard ways of overreaching: extending a flashlight test to every electronic device, blaming a temperature both groups shared, claiming a safety property nobody measured, and turning one measured ratio into a rule for all conditions. Five seconds against ten really is exactly twice as fast — once, at one temperature.

Two turn on the method itself. One experiment changes light and warmth together and so cannot say which produced the faster growth, however carefully everything else was arranged. The other asks what single change would let a student test a new factor, where the tempting wrong answer merely weighs the car: knowing a value is not the same as varying it.

The last two combine a pair of trials, and both are keyed to deliberately modest statements — "over this range", "in this species". Two points support a comparison. They do not support a rule, a requirement, or a prediction about anything beyond them.

Twelve flashcards carry the vocabulary and the reasoning: independent and dependent variables, controls, confounding, association against causation, and what a second experiment can establish that a first cannot.

Zestly is an independent study tool. It is not affiliated with ACT, Inc., which owns the test, and it is not an exam centre.

• Identify the variable deliberately changed and the variables held constant in a described experiment
• Tell a controlled variable from a measured outcome
• Say what a comparison group makes it possible to conclude, and what is lost without one
• Recognize that an experiment changing two things at once cannot credit either with the result
• Judge how far a result may be generalized beyond the conditions actually tested
• Reject a conclusion about a property nobody measured
• Design the single change to a method that would test a new factor
• State what two trials show together, in terms limited to the range they cover

Written for the research-summaries strand of the ACT science section, following the skills the test's own published description names: understanding experimental design, identifying variables and controls, comparing the results of related experiments, and judging which conclusions the data support. Because this is a text material with no figures, each experiment is described in words inside its own question and restated in full wherever it recurs. Metric units are used throughout, as in the real section. Every experiment, laboratory, student and sample is invented for this material, and nothing is taken from any ACT publication or released test.