Power Calculations – Quiz 4

1. Other things equal, what would happen to the power of the evaluation design if we increased the sample size by 10% (assuming an individual-level evaluation design, and an equally proportional increase in both the treatment group and the control group)?

2. Suppose that you have two evaluation design options: Either randomize 40 villages into T & C and survey 20 households per village, or randomize 20 villages into T & C and survey 40 households per village. Suppose that households within the same village have more similar outcomes than households in different villages. Other things equal, which design has greater statistical power?

3. Suppose that you are designing an evaluation and deciding between two outcomes to measure as your primary outcome of interest: household consumption and household income. Suppose that the intervention that you are evaluation will be randomized at the village-level. Household consumption has a higher within-village correlation than household income. Other things equal, which design has greater statistical power?