50 minutes. Self-marked. 12 marks. Write your answers, open the model answer, then tick each mark you earned honestly. Nothing is sent or stored.
Scenario (invented)
A news article reports a trial: 60 students were randomly assigned to write three good things nightly for two weeks (treatment) or to write neutral notes (control). Mean change in SWLS: treatment +2.4, control +0.6; standard deviation 6.0. The article says: “Gratitude lists boost life satisfaction, scientists prove.” The reported 95% interval for d is −0.21 to 0.81.
Task 1 (2 marks)
Calculate Cohen’s d.
Private: nothing you type here is sent or stored.
Task 2 (3 marks)
Interpret d and its interval in plain language, including the chance of superiority at that d.
Private: nothing you type here is sent or stored.
Task 3 (3 marks)
List three caveats about the article’s claim.
Private: nothing you type here is sent or stored.
Task 4 (4 marks)
Write the five-line evidence card for this single trial and grade it.
Private: nothing you type here is sent or stored.
Rubric and model answer
Task 1 (2 marks)
Difference in means = 2.4 − 0.6 = 1.8. d = 1.8 ÷ 6.0 = 0.30.
Task 2 (3 marks)
d = 0.30 is small to medium. The interval −0.21 to 0.81 includes zero, so no effect is compatible with the data, as is a large one. The chance of superiority at d = 0.30 is about 58%.
Task 3 (3 marks)
Any three of: “prove” is too strong; the interval includes zero; 60 students, one trial, no replication; two weeks only; students only; one outcome scale; effect may not last.
Task 4 (4 marks)
Claim: nightly gratitude lists raise life satisfaction. Mechanism: attention to good things. Best evidence: one randomised trial of 60 students, d = 0.30, interval including zero. Strength: Low (single small trial, no replication). Limits: short, narrow sample, one outcome.
Marks so far: 0 of 12