Statistical power is an indicator of the ability of a test of significance to “detect” a practical difference (e.g., between the averages of two products that are being compared). A low power typically means that the sample sizes in the study are too small. Without an analysis of statistical power, a conclusion of “non-significant” is rightfully questionable. Unless power is high, a study may be doomed to failure even before it is begun. This webinar provides thorough training in how to interpret and use the power-analysis outputted by text-book calculations or software programs modules (e.g., StatgraphicsCenturionXV).
Why You Should Attend:
Whenever a test of statistical significance is conducted with the hope that the result will be non-significant, the results may be unacceptable to a regulatory agency unless the test had an acceptable level of “power”. FDA typically requires a minimum of 80% power, and often requires 90% power. Calculation of power is so complicated that it typically must be done with a software program. Even so, the software program’s output can be misunderstood unless the user has a firm understanding of the basic concept of statistical power.
This webinar explains the basics, by using a t-test as an example. One of the very many possible formulas is then demonstrated, as well as 2 different software programs and their “Power Curves”.
Areas Covered in the Session :
Who Should Attend: