Final answer:
In a quasi-experiment, ethical and practical constraints often prevent random assignment, so researchers compare naturally occurring groups and control for confounding variables.
Step-by-step explanation:
Random assignment can be a challenge in a quasi-experiment as it often deals with situations where it is not ethical or practical to assign people to certain conditions purely at random. In a true experiment, random assignment is employed to ensure that any differences observed between experimental and control groups are due to the manipulation of the independent variable, not preexisting differences. Quasi-experimental designs, however, typically rely on non-randomized groups or naturally occurring groups, and researchers must then control for potential confounding variables. In the context described, it is suggested to compare individuals who choose to exercise with those who do not, rather than randomly assigning participants to engage or not engage in exercise. The ethical concerns of compelling individuals to change their behavior for an experiment often make quasi-experiments the more feasible choice, and researchers use statistical controls to handle lurking variables.