In epidemiology, a power of 80% (0.80) is generally considered acceptable. This means there is an 80% chance of detecting an effect if it exists. However, in some contexts, higher power (e.g., 90%) may be desired to ensure greater confidence in the study results.