When a test statistic is calculated from the sample statistic, such as sample proportion, it can be located in a probability distribution. This value of the test statistic demarcates an area under the curve from the rest of the area. This area at the tail of the distribution is the P-value, where P stands for probability. Assuming that the null hypothesis is true, there is always a chance of observing the calculated test statistic, or a value higher than that, in the critical region of the given distribution. A P-value provides the probability of getting that test statistic value in the critical region just by chance alone. So, when the P-value is observed to be smaller than a predetermined value—such as 0.05—we reject the null hypothesis, as it indicates that the observed outcome is highly unlikely and the evidence against the null hypothesis is stronger. The P-value can be calculated at the right tail, left tail, or both the tails depending on the hypothesis or the value of the test statistic.