SpawnLearn20 QUESTIONS · NO TIMER
Everyday Decisions
Compare numbers fairly, read evidence carefully, and make a confident claim only when the information supports it.
Try a round ↗Start with the denominator
A count alone rarely tells the whole story. Eighteen successful attempts can represent weaker performance than eight successes if the first group tried much more often. Ask 'out of how many?' before comparing. When combining groups, add successes and attempts separately. Averaging percentages without accounting for different group sizes can give a misleading result.
Look beyond the headline
A change from one complaint to two is a doubling, as is a change from a thousand to two thousand. The relative change is identical, but the scale differs. Percentage points describe subtraction between percentages; relative percent change divides that difference by the original rate. Keep both in view when a headline emphasizes one and omits the other.
Who had a chance to be counted?
A survey can contain many responses and still miss relevant people. Recruiting only frequent customers can favor their preferences. Publishing only positive reviews hides the experiences that were submitted but not shown. Ask who was invited, who responded, and which results were retained. A larger sample does not automatically repair a systematically selective process.
Separate a pattern from an explanation
When two things change together, consider other changes before attributing a cause. A website redesign during a sale does not isolate the redesign's effect. Random assignment can help make comparison groups more similar on average, though it does not guarantee identical groups or a positive result. The goal is to match the strength of a conclusion to the strength of the design.
Read the source material
Can you figure these out?
Ten questions. No timer. An explanation after every answer.
This guest round is not saved. Create an account to save future rounds and revisit the puzzle types you miss.