Significance testing is a statistical method used to determine if the observed results of a study are unlikely to have occurred under the null hypothesis, thus suggesting the presence of a real effect or difference. It involves calculating a p-value, which quantifies the probability of obtaining results at least as extreme as the observed data, assuming the null hypothesis is true, and comparing this to a predetermined significance level (alpha) to make a decision about the hypothesis.