Evaluating Statistical Claims
Evaluating generalizability and validity of statistical claims and study designs
Get a personalized JustLockedIn workbook paced to your test date.
Your workbook starts day one with a diagnostic that finds your blind spots, then builds targeted practice around the patterns actually costing you points.
You get a study description. The question asks: what can you actually conclude from this?
Why this matters
The SAT describes a study and asks what conclusion is valid. There are three specific patterns, each targeting a different reasoning mistake.
The three patterns
Pinpoint the Population
Identify who was actually studied. You can only generalize to the population the sample was drawn from — not everyone.
›Signal vs. Noise
Separate what matters (random assignment? random sampling?) from distracting details about the study. Design determines what conclusions are valid.
›Spot the Pitfalls
The wrong answers use subsets, supersets, or completely different populations than the one studied. They sound right but overreach.
The biggest trap: confusing correlation with causation. If the study didn't use random assignment to groups, you cannot conclude that one thing caused another — only that they're associated. The SAT tests this distinction constantly.