Fineuralab Learning Statistical literacy 中文

Transfer assessment: read numbers in new contexts

Five objective items and a written audit distinguish numbers, evidence and conclusions.

All scenarios are fictional. Answer first, then read explanations. Written work is not automatically graded.

1. Registration rises from 10% to 15%. What is precise? Up 5%, no baseline needed Up five percentage points, or 50% relative to baseline Up fifty percentage points

2. Five repairs take 1, 2, 2, 3 and 22 hours. What is wrong with “most take six hours”? The mean calculation is wrong The median must also be six Mean is six, median two; the mean does not describe most cases

3. A gym newsletter gets ten thousand exercise surveys. Does it represent all residents? A large sample does not automatically fix recruitment bias More than a thousand guarantees representation Decimal precision establishes representation

4. A test reports p=0.03. Which interpretation is invalid? Interpretation needs the null model and test conditions Effect size and uncertainty also matter The research hypothesis has a 97% probability of being true

5. Experienced users mostly receive a new service; beginners mostly receive the old one. New-service completion is higher. Is that causal proof? Yes, a higher rate is enough No; inspect baseline, assignment and measurement It proves the new service ineffective

Written audit

Before a redesign, 40 of 200 registered users clicked content; afterward, 60 of another 200 did. A promotion says the redesign increased user benefit by 50%. Examine arithmetic, denominator, comparability and conclusion scope.

My audit and remaining checks