An in-depth look at general population investigations of sleep disorders (especially insomnia) reveals many differences between them.

Inconsistent methodologies translate into inconsistent and incomparable results between studies.

Differences are seen in the types of questions and forms of questionnaires used to assess the presence and severity of sleep disorder symptoms and the impact symptoms have on daily functioning.

Finally, although some studies have suggested a connection between sleep disorders and an organic or psychiatric condition, studies with greater statistical power are greatly needed to confirm or reject such suspicions

Moreover, most of the epidemiological data we have on sleep disorders have come to us from surveys which consider the problem of sleep within a larger spectrum of problems, and thus important questions about sleep disorders are lacking, resulting in an incomplete picture of the problem.

With the benefit of the experiences from those who had gone before us and after years of planning and adjustments, we set out on the task at hand: an epidemiological survey of sleep disorders from general populations around the globe.

Our plan of attack for this daunting task required several fundamental items:

• First, in order to achieve thousands of interviews and record our interviewees' remarks "live", person to person, but without any medical involvement, we conducted telephone interviews using interviewers without medical training.

• Next, we took advantage of the many opportunities computer technology affords us today.

• A kind of artificial intelligence known as an expert system was designed and programmed by myself to uncover sleep disorders through the use of specific individualized questionnaires. Just as a medical doctor uses expertise to make diagnoses through a line of questions based upon the answers given to specific questions, so does this system. In addition this type of approach offered the advantage of improved data reliability.