What is Power?
- Power is the likelihood that an inferential statistical test will
correctly reject the null hypothesis
- Power is the likelihood of finding an effect when there is a real effect
What's the Problem with Low Power & Research?
- The power in past social science research has been estimated to average
approximately .60 -- i.e., it has only been able to detect real effects in
about 60% of studies
How Can I Increase Power?
- Increase the sample size (N)
- Increase the critical alpha (critical p level)
- Increase the effect size
|
Power Calculators
|