No one here denies that psych is an inexact science,bBut one thing is for certain: the field is rife with studies of one sort or another -- countless zillions of studies. For example, psychologists want to study some things about schizophrenia, its symptoms and/or causes and/or effective treatments, so they do an assiduous study, using real patients who've (already) been diagnosed as schizoid.

Psychologists and psychiatrists (for brevity I'll sometimes use the term "shrinks") usually diagnose their patients using a combination of elements. Primary of course are records forwarded by previous shrinks; but also of direct use are interviews of the patient, anecdotal contributions by those who have had significant contact with the patient, and observation of the patient's behavior in a controlled setting.

If a patient steadfastly refuses to release records or any information whatsoever about his previous psych consultations, then his current therapist will most likely offer to diagnose the patient by observing him/her over time in a controlled setting, possibly an institution.

So here finally is my inquiry: can anyone cite a study that has ever been done to determine the efficacy of psychiatric diagnosis?? I am talking about a double blind controlled experiment. Let's say you took a couple dozen various individuals, a small undisclosed number of whom have been diagnosed as suffering from one or another psych malady, and put all these individuals into a controlled setting where they can be observed and interviewed (maybe videotaped?). And then let's say ten different shrinks are independently shown the videotapes and asked to identify which of the individuals are sick, and in what way.

I'm sorry if my wording of all this is imprecise, but I am sure you get the idea. In the patient interviews, certain questions would have to be off limits, such as (obviously) if they've ever seen a shrink or been diagnosed. More interview content constraints would have to be imposed as well, obviously. The whole idea behind such a study is whether or not shrinks, working independently from one another, could arrive at matching conclusions about which of the study subjects is psychologically afflicted, and in what manner, based solely on observations and interviews. Of course, the study subjects would (in one variant of the study anyway) be permitted or even encouraged to employ some deceptions.

I would be extremely interested to read all about the results of such a carefully blinded study and see what if any concurrence manifests among the shrinks.

Obviously, this study would have vital implications to Law and Ethics, since it would demonstrate the degree of confidence one can presume in psychiatric diagnosis, when cheat sheets are unavailable... "cheat sheet" meaning the patient's prior records.