The Scientist and the Practitioner: Some thoughts on A Vexed Relationship

The first thing I ever knew about clinical psychology was that it was based on the "scientist-practitioner" model. This paradigm, dating from a conference in Boulder Colorado in the 1940s (just as clinical psychology was taking off after the second world war), has largley defined the profession on both sides of the Atlantic ever since.

However, despite the fact that the science of psychology has expanded wildly since 1949, the scientist practitioner model seems ever less central to the profession. There are instances of outright disregard for the "scientist" part evident in many quarters and I find myself in interesting debates with psychologists and other therapists, trying to figure out how to define the role of "evidence" in that cumbersome phrase "evidence based practice". For many I detect a current of hostility to thinking in data and a general preference for using common sense and seeing the person with whom you are working. Data, by some accounts, is the kind of thing upon which we can become "hung up". Perhaps people are wary of becoming cold hearted Spocks:

"I'm Listening"

Although I have previously been scornful of the woollier arguments used against evidence in psychology, I have sympathy for some of what people are (I think) trying to express in these reservations. This post is an attempt to persuade them there is less disagreement than they fear.

Basic Problems:

Let's put our hands up right away and get a few things straight about the science of therapy. First, what gets called "evidence based practice" is not necessarily always (or even most of the time) living up to the lofty ideal of the name. For some institutional bodies, the existence of plausible seeming figures in some journal is good enough to plough ahead and recommend a therapy. I have even seen people make quite important decisions on the basis of a single pie chart (of unknown provenance) in a promotional brochure:

Even Data Can be Meaningless

Equally, much research is compromised by financial interest and driven by large pharmaceutical companies. Furthermore, the terms under which research is conducted are largely defined by political considerations of what is in vogue or popular. CBT gets more attention than any other modality, especially ones that don't sound snazzy or mysterious (like ordinary "befriending" or "supportive therapy").

Empathy and Therapeutic Skill:

There are definite limits for the role of science in the practice of therapy. The skills and considerations of an effective therapist can probably be determined by data, but perhaps their cultivation is a separate thing. I am sure there is no contradiction between being an empathic, kind and effective therapist and being statistically competent to assess efficacy, but I am aware of no reason to believe that proficiency in one automatically helps with the other. Whatever it is we do to cultivate kindness and empathy (an empirical question), it seems pretty clear we should do it.

The Role of Values and Outcomes:

The question of what kind of society we wish to have cannot always be answered empirically. A neat example of this is in a debate I recently had with a friend. In New York, unlike London, people have to pay to get into museums. In a basic way this doesn't feel right to me and I was trying to articulate why. "Societies just seem better when their art and culture is accessible to anyone" I argued. "Can you prove that?" asked the friend, and of course I couldn't. There may exist some quantitative indications that free cultural activities are good for people, but I doubt they are very robust, and in any case I am not interested in them. Even if you couldn't show conclusively that free access to culture improved people's lives, I would still maintain it was a social good. This is because it is not a question of tangible effects, but of what kind of society I want to live in.

To transpose this onto the field of mental health and social care; at least some proportion of what is at stake in the debates cannot be settled by data. Walking onto a mental health ward for the first time, I was struck by a sense of how cold and cruel it seemed. Apparently abandoned residents walked about in distress and staff members callously (so it seemed to me) bossed and condescended to them. Leaving the place behind felt like a palpable relief. Who knows how you would begin to quantify what it is like to live in such places; what kind of impact it would have on your sense of self. This is not to say that we shouldn't try, nor that we can ignore good quality data, but we can't depend on quantitative data to know everything we feel to be worth knowing. Sometimes the feelings that philosophers call "intuitions" are worth listening to. Debates outside of science, about what we should value are worth having too.

Rapprochement:

However, none of these readily acknowledged limitations is straightforwardly an argument against the centrality of the scientist practitioner position. I see a definite tendency to martial the limitations of science-based practice and attempt to assemble them into a case against data, but that way danger lies. One line of attack is to identify a scientific persuasion as a kind of arrogance. The logic apparently being that the scientifically oriented are vulnerable to using data to somehow over-ride the immediate experience of the service user, perhaps by steadfastly maintaining that they continue to do something (take a pill, undergo a form of therapy) when it is not working for the individual. This is indeed a risk of following evidence-based recommendations blindly, but it would in fact be a deeply unscientific thing to do. Reasoning from samples to individuals is probabilistic, and even if an intervention worked for 95% of cases, there are still 5% for whom it won't. The rational scientist-practitioner treats every case as a new instance of reality and pays attention to what is and is not working. This is what Jacqueline Persons (here) calls "Treatment as Experiment".  People who say that therapy is an "art" or that clinicians need to "be confident enough not to need to know all the answers" can, I hope, see a direct parallel with what they are proposing. 

Therapy, like most complex human behaviour, can probably be described as an art, but that doesn't mean it can ignore science. Is architecture an art? Plausibly yes, but if architects ignore the principles of engineering and physics, their buildings will kill people.