When you are going to make a prediction about a person or an event based on a set of data, there are two possible modes of data combination available. One is the clinical method which relies on human judgment, experience, informal contemplation etc. to reach the final prediction. The other is the statistical method which involves a formal, algorithmic and objective procedure (often an equation) to make the prediction.
Since the 1950s there has been considerable controversy surrounding the “clinical vs. statistical debate”. Even though most studies comparing clinical vs. statistical prediction point in the direction superiority of the statistical approach, not many have taken the consequences. This thesis reviews the literature and theories in the field in the tradition following Meehl (1954) and in doing so the thesis paints a picture of an ideal incorporation of the principles of the debate as opposed to the lack of incorporation of the same principles in the world today. Further, the thesis examines possible reasons why the research have had such a small impact.
Possible reasons for the small impact include a range of objections expressed by experts themselves, lack of education in statistics and probability theory, and lack of knowledge about the clinical vs. statistical controversy.