1. the likelihood of incorrectly rejecting a statement or hypothesis concerning a characteristic of a population. More specifically, it is the probability of incorrectly rejecting a true null hypothesis (i.e., committing a Type I error) in research. Although the value of alpha is chosen by the investigator based on what is deemed acceptable for a particular study, alpha values of .05 and .01 are commonly used.

2. a measure of reliability for a set of responses to a test or measure. See Cronbach’s alpha.