Statistics HELP!!!

Question

You plan to conduct an experiment to test the effectiveness of Sleepeze, a new drug that is supposed to reduce insomnia. You will use a sample of subjects that are treated with the drug and another sample of subjects that are given a placebo.

1.) What is "blinding" and how might it be used in this experiment?

2.) Describe how you would use replication in this experiment.

I am hoping someone could just explain blinding and replication in the context of this problem. I'm new to statistics and not really sure what they are.

 

 

Answer 1

Tilda: If subjects in a group KNOW they are getting a placebo, it does not provide a valid comparison. Also, there may be opposition from subjects on ethical grounds, unless a signed consent is obtained earlier. To avoid this drawback, blinding is used. A patient does not know whether or not s/he is the treatment or the placebo. They are administered identical looking containers. Many times the person who administers also does not know ( double blinding ) who gets what. Thus blinding eliminates that kind of bias. In the context of the problem, there are many designs possible. In a "matched pair" design N pairs of patients of same age-group, same gender, same existing conditions etc. are chose. Their sleep hours are recorded for m days. Then randomly one from the pair is given the placebo, w/o being told so. The administerer does not know either. The other person in the pair gets Sleepeze. Again their sleep hours are recorded for n days (m may equal n). Here replication used to increase precision of the experiment. If you need further assistance I can talk to you on the phone, if first you file your payment info with Wyzant. Dr, g.