The probability of obtaining the realized test statistic value, or a more extreme value, is the proportion of L sampled test statistic values equal to or more extreme than the realized value, where L is usually large (Johnston, Berry, & Mielke, 2007)