Comparing experiments on ACT Science: identifying the one design difference between two related experiments and using paired results to attribute an effect to that difference.

What this topic is asking

Many Research Summaries passages run two or more related experiments for a reason: so the test can ask you to compare them. A comparison question turns on one idea: find the single design difference between the experiments, then read their results side by side to see what that difference did. Done carefully, these questions are some of the most reliable points on the section.

Step one: find the single difference

When a passage presents Experiment 2 as a near-repeat of Experiment 1, the experiments are usually identical except for one feature. That feature is the point of the comparison. Look for the sentence that says Experiment 2 was done "the same as Experiment 1, except ..." or scan the methods for the one thing that changed: a different concentration, a higher temperature range, an added catalyst, a new species, an extra step.

Step two: compare like with like

Once you know the difference, compare the results under matched conditions. If Experiment 1 and Experiment 2 both test a range of temperatures, compare them at the same temperature, not at different ones. Reading the rate at 30 degrees in both experiments isolates the effect of the changed variable, because temperature is held equal and only the changed feature differs.

This is the same controlled-comparison logic as isolating a variable in a table (reading tables and multi-variable data) and as the design of controls (variables, controls, and experimental design).

Step three: attribute the effect

With one difference and a matched comparison, any change in the result is the effect of the changed variable. If the rate doubles from Experiment 1 to Experiment 2 at every matched temperature, and the only difference is the doubled acid concentration, then doubling the acid concentration increased the rate. The matched comparison licenses the causal claim that a single uncontrolled difference would not.

Knowing the limits of a comparison

A clean comparison supports a specific conclusion, not a sweeping one. If two experiments differ only in acid concentration and were run at one temperature, the comparison shows the effect of concentration at that temperature, not at all temperatures. The ACT often offers an over-broad answer ("acid concentration always doubles the rate") that the matched data do not support. Pick the conclusion the comparison actually licenses, a judgement that overlaps with evaluating models and inferences.

Try this

Q1. Experiment 2 repeats Experiment 1 but uses a catalyst. Both are run over the same temperatures. How should you compare them to find the catalyst's effect? [2 points]

• Cue. Compare the results at the same temperatures in each experiment; because the catalyst is the only difference, any change in the matched results is its effect.

Q2. Two experiments differ only in light intensity, both run at 20 degrees Celsius. The plants grew taller at higher intensity. What can and cannot you conclude? [2 points]

• Cue. You can conclude that higher light intensity increased growth at 20 degrees; you cannot conclude it does so at other temperatures, which were not tested.