Why is the random assignment of individuals to experimental and control groups important for an experiment?

Without random assignment there is a danger of systematic error

• or bias - entering into the results. Statistical theory depends

on the errors being random and independent error and that is no longer the case without random assignment.

In fact, statistical experiments are often "double-blind": even the observer does not know which individual is in which group. This is to prevent unconscious or subconscious messages to affect the outcome (placebo effects).