The process of determining the sample size required to achieve certain statistical power or precision of estimation.