|
Field
Power calculation: Minimum Detectable Effect Size for Main Outcomes
|
Before
We use a one-tailed test, alpha = 0.05, beta (power) = 0.2 to calculate our sample size. 1) For a general dictator game, under the situation that a dictator needs to allocate 100 units between herself and the recipient, previous studies with similar subjects have shown that the variability is approximately normally distributed with a standard deviation of 25, the minimal relevant difference equals 10. The required sample size for each group is 78 (Gruener, 2019). 2) For a design that outcome is a noisy function of effort (Brownback Kuhn, 2019), if the minimal relevant difference is 0.07 with a standard deviation of 0.17. The required sample size for each group is 73. Considering information from these two literatures, we will have 80 samples each treatment.
|
After
We use a one-tailed test, alpha = 0.05, beta (power) = 0.2 to calculate our sample size. 1) For a general dictator game, under the situation that a dictator needs to allocate 100 units between herself and the recipient, previous studies with similar subjects have shown that the variability is approximately normally distributed with a standard deviation of 25, the minimal relevant difference equals 10. The required sample size for each group is 78 (Gruener, 2019). 2) For a dictator game with real effort design (Heinz et al., 2012), if the minimal relevant difference is 10 with a standard deviation of 24. The required sample size for each group is 71. Considering information from these two literatures, we will have 80 samples each treatment.
|