To ensure the researcher doesn't know who is treated.

To measure the psychological effect of receiving any treatment.

To increase the standard deviation of the results.

To force the participants to drop out of the study.