A therapist wants to study the effects of yoga and meditation on stress relief. She has 60 volunteers who experience varying levels of stress. Half of the participants will be assigned to practice yoga for one month and the other half will practice meditation. Before the experiment begins, all of the participants will be asked to rate their stress levels on a scale from 0 to 10, with 0 representing "no stress” and 10 representing "highest level of stress.” The therapist believes that current stress levels could affect the results; therefore, she has the subjects participate in both treatments. For each subject, a coin is flipped. If it lands on heads, then the subject will practice yoga for the first half of the month and meditation for the last half of the month. If it lands on tails, the subject will practice meditation for the first half of the month and yoga for the last half. Stress levels are measured before switching treatments and again at the end of the month. The differences in stress levels will be compared. Does this describe a matched pairs design?

Yes, each subject is randomly assigned both treatments.

Yes, each subject was asked to rate their stress level at the beginning of the experiment.

Yes, but the effects of the first treatment may affect the results of the second treatment.

No, the subjects’ stress levels before the beginning of the experiment may affect the results.