A p-value of 0.05 or lower is generally considered statistically significant. P-value can serve as an alternative to—or in addition to—preselected confidence levels for hypothesis testing.

State a null hypothesis: that is, there is no effect. Calculate the p value, which is the probability of getting results like ours – if the null hypothesis is true.