When I speak of pseudoscience, I mean a mode of study that only seems to follow a scientific method of proof but in reality doesn’t and when I speak of foundation I mean concrete and consistent set of axioms that may me be based on underlying more fundamental axioms or branches of logical argument from another such axiom system. This is very recursive definition but for now it would suffice,I guess!
So this post is about psychiatry and how messed up it has become. The fact is that whole science of medicine is screwed (I would write about it in detail later) but particularly psychiatry has crossed all the limits of rationality.
In psychiatry the diagnosis of disease is based on two fundamental criteria, namely. ICD 10, or international classification of disease 10th review, which is WHO’ s effort to classify mental disorders and DSM IV-TR or Diagnostic and statistical manual IVth edition -Text Review. ICD these days is used more commonly to classify subjects in research studies and DSM is usually used in routine clinical practice. Quite contrary to what name suggests, DSM (Diagnostic and statistical Manual) has nothing to do with statistics. What happens is that handful of very experienced Psychiatrists decide what signs and symptoms to include in a given syndrome or disease and with mutual debate decide upon the criterion of diagnosis. Off course there are statistical studies that are carried out, but they are poorly designed surveys and very few of the studies can, in true statistical sense, prove hypotheses. Very few people in science realize this but the true essence or goal of a statistical study is to prepare grounds for formal models (like mathematical models) that predict something. Prediction or implication of a theory is also important, without it even a great theory can potentially fail. A good example of this is the famous String Theory in Physics. But prediction is the second step, first comes the coherence of statistical studies to make a formal model that describes a system, well. Another thing is generalizability of a study, which means capability of a study to reflect upon the entire population. Studies in Psychiatry lacks both of coherence and generalizability, so it is impossible to build upon them. But Psychiatrists are doctors and scientific method is not taught in medical schools, hence they make this big blunder when it comes to extending their own field of study. What they do is they combine very few, un-generalized and non-coherent studies and mostly from their experience they construct a model of disease. Now what would a sensible person think, would happen to such a model, it would fail off-course and boy does it fail!!
Because of this psychiatry has become something that is neither valid nor reliable. Validity in science means that capability to measure what is being measured and reliability means that what is being measure is being measured consistently. Now imagine you went to a market to buy an instrument. Lets take it to be a weighing scale and you get to know that
the scale is invalid or it has an error and you cannot measure your weight precisely would you buy it? Well I might buy it because I am often more interested in managing my weight and at-least I would know whether I am loosing or gaining weight if it reports the weight by exactly the same error (that is if it is in-valid only). But imagine a scale giving not only erroneous but inconsistent results (that is unreliable) ..in that case even I wont buy it ! Same is the case with Psychiatry. All I wanted to demonstrate from my parable of weighing scale is that absence or presence of such systems of diagnosis makes no difference. Because at all times you will have a degree of uncertainty and until and unless you don’t know how large is the magnitude of that uncertainty (in psychiatry you’ll never know), you will have no advantage in knowledge.
Becuase medicine is based on statistics (or claims to at least ) and the proof of whether a treatment is working or not depends on measuring the effect before and after administration of a treatment in a population of patients. It is all about measurement! Now lets go back to our weighing scale parable. Because my scale is neither relaible nor valid it would give me a string of readings if I weigh myself. for example it might give me 104, 85.6, 62, 75, 89 (kilos) and now suppose I get to know about a break through drug that burns fat. I took and weighed myself again now the readings I get, after series of measurements is 22.3, 44.5, 102.7, 98.9, 55.3. Now can you guess whether this drug is working? you cannot **
This all does not imply that we are helpless and nothing can be done. There are methodologies that have been devised by great mathematicians and computer scientists, that can help us in making psychiatry a science with solid foundations. I will discuss such methodologies shortly in my next post.
** If you are thinking that you may take the mean values of both strings, then you are assuming that the deviation of error will depend on the proximity of original measurement (the one you tend to measure i.e your weight). This assumption cannot be made until and unless you have done previous experiments that reveal such property.
Über-Coder 