Planning research studies: an alternative to power analysis.

Hypothesis testing is a cumbersome method of statistical analysis and sometimes leads to misinterpretation of results. Whenever possible, it is preferable to design studies and analyze data using estimation procedures. For ANOVA, correlation analysis, regression, and other standard statistical methods, it is possible, and no more difficult, to use estimation procedures.