• α ("Alpha") is the probability of concluding that there is a difference between the groups studied, but in reality there is no difference (also known as making a "type I error").
• α is usually set a-priori to be 0.05.
• A finding is considered to be statistically significant if the p-value obtained is < 0.05.

• Simplistically, α is the chance of making a type I error. 
• The smaller the α value the lower the risk of making a type I error.  Note: doing this also usually means the investigator will need a larger sample size to a difference if it exists.

• The alpha value and sample size are the two most common contributors to the power of a study.
• Setting the alpha to be smaller (e.g., going from an α < 0.05 to an α < 0.025) prior to starting a study will make it harder for investigators to find a difference if in fact a difference is to be found.
  
• However, if a difference is found when doing this, the risk of making a type I error is lower.

• P-value
• Beta (β)
• Power (1 - β)

Editors & Reviewers:

• Anthony J. Busti, MD, PharmD, FNLA, FAHA
  
• Karolina DeAugustinis, MD

Last Reviewed: July 2015