Power

Power

• As mentioned in previous modules, post-hoc power (calculated after conducting the analysis) is meaningless.
  • Power is a direct function of sample size.
• The best way to use power is to specify it and the effect size you would like to attain, and use those to calculate the number of subjects/participants needed to achieve those goals.
  • This is called a-priori power calculation.
• A-priori power calculations (i.e. calculating the necessary sample size) should always be done when planning a study.
  • A-priori calculations should be done prior to data collection.

G-power 3

• The best way to calculate a-priori power/sample size is to use G-power 3, which is available for free at the link below.
  http://www.psycho.uni-duesseldorf.de/abteilungen/aap/gpower3/
• At the G-power 3 web site, you will find links to the User Manual for G-power 3 which are very easy to follow and guild you through the process of carrying out the necessary calculations for a variety of analyses.
• Also at the G-power site, a Literature section is available which contains a references list related to power and sample size calculation.