Power greater than 0.50 are traditional in social
sciences. As reported, a statistical test
must have a power of 0.80, and the probability of a
type II error of 0.20. By these criteria, we lost one in
five genuine effects. However, reported that in the social sciences, power
usually ranges from 20% to 50%. Therefore, these
investigators recommend using a power of 0.50 in
future similar research, where the intervention has
minimal known harmful effects.