Skip to content
SAT · Problem Solving and Data Analysis

Causation

Causation appears in 4 reviewed SAT questions and sits in Problem Solving and Data Analysis. Below are worked examples showing exactly how the exam tests it, with the answer marked and every distractor explained.

How Causation is tested

Example 1Harder
The following text is original practice text. A psychologist is reviewing studies on social media use and depressive symptoms in teenagers. She notes that most correlational studies find a positive association between heavy social media use and higher rates of depression, but argues that this correlation does not establish causation. She plans to propose a longitudinal study that tracks the same adolescents over several years to better understand the direction of the relationship. Which finding, if true, would provide the strongest evidence that social media use causes depressive symptoms rather than the reverse?
  1. ATeenagers who significantly reduced their social media use over six months showed corresponding decreases in self-reported depressive symptoms.Correct
  2. BTeenagers who report higher rates of depression also report spending more hours per day on social media than teenagers with lower depression rates.
  3. CDepression rates among teenagers have risen steadily over the same period during which social media platforms became widely available.
  4. DTeenagers with depression report using social media primarily to seek social connection, while those without depression use it mainly for entertainment.

Why A is right

To show that social media use causes depression (rather than depressed teens seeking more social media), we need evidence that reducing social media use leads to improved mood. This experimental-style evidence demonstrates a directional, causal relationship.
Why the other 3 choices miss
B
Incorrect because this describes a correlation between social media use and depression, not causation—it cannot determine whether social media causes depression or depression increases social media use.
C
Incorrect because a historical correlation between the rise of social media and rising depression rates does not establish causation; many other factors changed over the same period.
D
Incorrect because differences in how depressed and non-depressed teens use social media do not establish that social media causes depression.
Example 2Harder
A researcher performs a controlled experiment and randomly assigns 50 volunteers to receive a new medication and 50 to receive a placebo. Which statement is most valid based on this study design?
  1. ACausal conclusions about the medication's effect are supported.Correct
  2. BThe study cannot be generalized to any population.
  3. CThe placebo group's results are irrelevant to the conclusion.
  4. DAny observed difference in outcomes cannot be attributed to the medication.

Why A is right

A randomized controlled experiment allows causal conclusions; random assignment controls for confounding variables. The data or units point to this result when the values are interpreted correctly.
Why the other 3 choices miss
B
Incorrect: Random assignment allows causal inference; generalizability depends on how volunteers were recruited.
C
Incorrect: The control group (placebo) is essential for comparison; without it, you cannot determine the treatment's effect.
D
Incorrect: Randomized controlled experiments are specifically designed to attribute differences to the treatment.
Example 3Medium
To study the effectiveness of a new teaching method, a school randomly assigns half the students in each class to receive the new method and half to receive the traditional method. At the end of the year, test scores are compared. Which conclusion is most appropriate?
  1. AThe study can determine causation because of random assignment.Correct
  2. BThe study is not valid because students interact with each other.
  3. CThe results cannot be applied to students in other schools.
  4. DThe traditional method is automatically better since it is established.

Why A is right

Random assignment within the experiment allows causal conclusions about the teaching method's effect. The data or units point to this result when the values are interpreted correctly.
Why the other 3 choices miss
B
Incorrect: While student interaction is a potential confound, random assignment minimizes its effect; the study remains valid.
C
Incorrect: Generalizability depends on whether the sample is representative; the study may or may not generalize.
D
Incorrect: Statistical evidence from the study, not tradition, determines which method is more effective.

Studied alongside

Other concepts from Problem Solving and Data Analysis.