| Collect volunteers for a study. Divide them into two random groups. In the first group, each volunteer has a one-on-one "consultation" with a professional astrologer. In the second group, each volunteer has a one-on-one "consultation" with an actor pretending to be an astrologer, but who is really making up convincing-sounding advice at random. (However, volunteers don't know what group they're in, so all volunteers think they are talking to an "astrologer".) A week later, ask the volunteers to give a rating to the "astrologer" that they talked to, based on whether their advice was good and their predictions accurate. See if the real astrologer got higher ratings than the actor. |
| Find a group of people who consult astrologers regularly (for example, you could post a notice in an astrologer's office, with the astrologer's permission, asking customers to sign up for a study). Ask those people, on a scale of 1 to 10, whether they believe that their astrologer gives good advice and makes accurate predictions in general. The higher the ratings, the greater the likelihood that astrology works. |