Selection bias occurs when the participants selected for a study are not representative of the target population, leading to skewed results. In the context of cancer research, this can significantly affect the validity of findings and the effectiveness of interventions.