Sample Size Planning for Statistical Power and Accuracy in Parameter Estimation

Excerpts

1. Sample size planning is important to enhance cumulative knowledge in the discipline as well as for the individual researcher.
2. Sample size planning can be based on a goal of achieving adequate statistical power, or accurate parameter estimates, or both.
3. Researchers are actively involved in developing methods for sample size planning, especially for complex designs and analyses.
4. Sample sizes necessary to achieve accurate parameter estimates will often be larger than sample sizes necessary to detect even a small effect.
5. Sample sizes necessary to obtain accurate parameter estimates or power to detect small effects may often require resources prohibitive to the individual researcher, thus suggesting the desirability of study registries accompanied by meta-analytic methods.

Reference

Scott E Maxwell, Ken Kelley, Joseph R Rausch “Sample Size Planning for Statistical Power and Accuracy in Parameter Estimation” (2008) DOI: 10.1146/annurev.psych.59.103006.093735

@Article{maxwell2008,
  title = {Sample Size Planning for Statistical Power and Accuracy in Parameter Estimation},
  volume = {59},
  issn = {1545-2085},
  url = {http://dx.doi.org/10.1146/annurev.psych.59.103006.093735},
  doi = {10.1146/annurev.psych.59.103006.093735},
  number = {1},
  journal = {Annual Review of Psychology},
  publisher = {Annual Reviews},
  author = {Maxwell, Scott E and Kelley, Ken and Rausch, Joseph R},
  year = {2008},
  month = {jan},
  pages = {537–563}
}