Posted on May 26, 2010.
Problem with statistics ...? The doctor who has practiced for over 20 years estimates that goes "cold turkey" is more effective than nicotine patches and other methods of smoking cessation help smokers quit. The doctor gives as proof of his observation of patients in his practice.
a. Explain the limits of such anecdotal evidence.
b. Mention of a hidden variable that could affect the medical association was observed.
smoking cold turkey, he perhaps sees everything he does in a hospital where they had no other choice. it would be more likely to see this kind of quit.
there is no treatment and control data current
it has no data on the number of people do not go to his practice that have tried to quit, that the method and the success of failure.
it can not account for age, sex, smoking years or number of cigarettes smoked before quitting, all the factors of success regardless of the method.
A claim like this must be supported by a study where participants are divided into groups. Ideally, the study should be double blinded (neither the patient nor the doctor know who is in which group). The limitations of this study are the sampling method used, the measurement of efficiency and distribution into distinct groups.
A variable can be hidden own physician involvement in the process. He may have given more attention to some patients than others. Therefore, double-blind studies are done.
In addition, the doctor may have his own prejudices!