P-value is the level of marginal significance within a statistical hypothesis test, representing the probability of the occurrence of a given event.

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.

The test depends on an extrinsic hypothesis, because it requires theoretical expected values to be calculated.

Consider n normal populations. Random samples of equal size m are taken from each population. It is desired to use the Bartlett procedure to test at level α the hypothesis of equal variances. A method ...