Define and frame
State H0 and H1 for a question about differences between two means, noting one-sided versus two-sided alternatives.
Scenario: A study reports p=0.04; another runs 20 independent tests. Carefully explain what the numbers can and cannot say.
Use the fictional dataset if the question refers to it. Always show your reasoning; an exact number without context is not enough. The four written tasks are teacher-assessed. The self-check below is formative and does not affect the 20-mark assignment.
Written problem set
Calculate or apply
Translate p=0.04 into a conditional probability statement under H0 and test assumptions.
Evaluate the design
Under 20 independent tests with true nulls and α=0.05 each, what is the expected number of false positives?
Interpret and communicate
Why is it problematic to conclude “there is no effect” from a non-significant result? Refer to power and intervals.
Standard scoring guide
Each written task is worth 5 points: correct key idea (2), justified application or computation (2), and a clear statement with limits (1). This original rubric is not official UCL assessment.
Formative self-check · 4 questions
Choose one answer for each item, then check the feedback. Answers and explanations are available on this public study page; the separate written-task instructor key is private.