Research Summaries

• Identify the independent variable, dependent variable, and controlled variables in an experiment
• Understand the purpose of a control group
• Compare results across multiple experiments within a passage
• Predict results for new experimental conditions based on observed trends
• Finish on one ACT-hard question and say what made it hard

Research Summary passages describe one or more experiments, including their design, procedures, and results. The independent variable is what the scientist changes; the dependent variable is what is measured; controlled variables are kept constant. A control (or baseline) trial is the one the others are compared against, often using a zero, normal, or untreated value. This does not always mean a biology-style 'untreated group': a control might be the reaction run with no catalyst, the sample at room temperature, or the measurement with the voltage set to zero. When a passage describes multiple experiments, the ACT often asks how they differ. Typically, each experiment changes one variable while holding others constant. You may also be asked to predict the result of a trial that was not run: between two tested values, read the trend between them (interpolate); beyond the tested range, extend the trend (extrapolate), and check whether the data were leveling off. Research Summaries passages are the most common of the three formats, so their design questions are worth mastering. To test whether a factor matters, change only that factor and hold everything else constant. Repeating trials or using larger samples makes results more reliable but does not change what is being tested. A result supports a hypothesis only if every relevant condition moves in the predicted direction, so check each row, not just the overall trend. When a question asks why a step was included, the answer is usually that it controls a variable that could otherwise explain the result. Before the questions, restate the research question in your own words (what effect of what on what), since the design questions all turn on it. The best-sounding proposed design is often a trap because it changes two things at once. When a passage has three or more experiments, sort the questions first into look-ups in one experiment and comparisons across experiments, and check which experiment each value comes from. Two studies that differ in more than one way cannot show the effect of either difference. The ACT-hard version: a Yes/No question where every reason quotes real numbers from the passage and only one reason both matches the data and answers the claim, or a new trial that has to be read from two experiments at once (find the matching condition in one experiment, then carry it to the other).

• Before answering questions, list the variable changed and the variable measured for each experiment.
• If a question asks 'which experiment...,' compare the experimental designs side by side.
• For prediction questions, find the trend in the data and extend it logically. Do not guess.