Power Analysis Reporting
The power of a statistical test is a crucial property in making inferences, especially in experiments, since statistical power is determined in part by the sample size, or the numbers of trials run in the experiment. Standard practices require an assessment of the statistical power, and perhaps an adjustment to the research design, prior to the experiment. One should estimate the size of the experimental – control difference that represents a meaningful, substantively important difference. This is the difference one wants to declare as 'statistically significant'.
The level of significance needs to be set, the size of the type 1 error (probability of falsely rejecting the null hypothesis of no difference). The power of the test should be specified (the probability of correctly rejecting the null hypothesis if the true differences between the control and experimental results is at least as large as previously specified. One then needs to specify the type of statistical test that will be used. Working backward from these parameters, it is possible to calculate the number of subjects needed for an experiment, or the size of a sample. Standard practices require this information to be part of the analysis presented.
|
 |
|
|
