Aristotle in conference with modern psychologists
All posts

The Golden Mean: Balancing the Scale of Evidence-Based Practice

Virtues lie between two opposing excesses. So argued Aristotle in his Nicomachean ethics, over 2300 years ago. So, courage can be understood as the mean between recklessness and cowardice. Temperance falls between self-indulgence and insensibility. Pride is found between vanity and undue humility.

One of the brilliant features of this analytic framework is that it breaks us out of our habitual way of seeing virtues as simply ‘more’ of something. Courage is good, and so one should surely be as brave as possible. But no, says Aristotle. Because if we push it too far courage becomes recklessness.

In a conversation with Dr Shane McLoughlin, a good friend and colleague, this idea of Aristotle’s Golden Mean came up, followed by a discussion of the concept of evidence-based practice in applied psychology. It was perhaps the proximity in time that made me realise how relevant each is to the other.

We are inclined to think of evidence-based practice as a virtue — a mono-polar virtue. More evidence-based practice is good. Requiring more evidence is good. What if this is just the sort of unsophisticated thinking Aristotle was warning us against?

The danger of requiring too little evidence in psychological practice is clear: it opens the door to charlatanism. Practices based on insufficient evidence can lead to ineffective or harmful interventions. For instance, treatments that have not been rigorously tested may fail to provide relief, or worse, they might cause psychological, emotional, or financial harm to clients. In coaching psychology, ineffective intervention could hold people back in their professional lives, resulting in considerable lost earnings, not to mention undermining the credibility of our profession and eroding public trust in psychological services.

What is often less obvious is that demanding excessive evidence before implementing any intervention can also be very detrimental. Human beings are complicated. Actually, to be more technically precise, we’re complex. There are non-linear interactions between different features of human behaviour, with adaptation and feedback both within and between individual humans. There is substantial individual variation between people, a lot of this driven by the different life experiences we’ve lived up to this point. Even our best and most well-replicated models to predict behaviour in a given situation typically leave about half the variance unexplained. We will never have anything remotely like a dead-cert as a behaviour change technique. We have techniques that work well for many people in many circumstances. But we will never have a guarantee that a selected intervention will help this person for this issue. Most of the evidence informing psychological interventions is derived from group-based studies where we examine the average change in a group based on exposure to the intervention.

What’s more, these group-design studies are set up with a specific outcome or set of outcomes in mind. If a client turns up wanting to achieve some specific goal, there simply might not be a study that (a) includes participants very much like this client in temperament, life experiences, and demographics and (b) where the outcomes measured are precisely the same as the one the client desires. In fact, I would say we almost never have this type of perfectly relevant evidence.

When people want a psychological intervention — coaching perhaps to reach some professional goal or to lose weight — if those who care about evidence-based practice say ‘I can’t help you’ because we believe we have too little evidence, most clients will not just give up. Most of them seek help elsewhere. And guess who they’ll end up seeing? By definition, clients will end up seeing someone who doesn’t care about evidence-based practice.

If they can’t or don’t want to find someone with less caution, then perhaps they’ll leave the itch unscratched, but even that has its cost — goals unachieved, unhealthy weight still carried, life satisfaction lower than it might have been.

I would argue that Aristotle’s Golden Mean applies just as well to Evidence-based Practice:

The virtue of evidence-based practice lies at the mean between charlatanism and over-caution.

But if we are to offer help to our clients based on less psychological evidence than we would like, how can we claim to be doing so ethically? The formula is already woven through the fabric of our culture and indeed our practice: informed consent.

Financial advisors are always necessarily working from very inadequate evidence. Economic and political events over coming months and years will affect the markets in ways they cannot predict with anything even remotely approaching certainty. And what do you hear at the end of every advert for a financial product? “Investments are subject to market conditions and can fluctuate in value, both upwards and downwards, and you may not get back the original amount invested. Past performance is not indicative of future results.”

The British Psychological Society’s Practice Guidelines have some excellent advice on gaining informed consent, including that:

  • The process of obtaining consent is dynamic and must be revisited whenever significant changes occur in the intervention or the client’s willingness to continue.
  • The influence of the psychologist’s authority and the client’s desire for help can complicate the consent process, potentially leading to compliance rather than genuine agreement.
  • It is crucial that consent discussions are clear and ongoing, allowing clients to actively participate and ask questions to facilitate informed decision-making.

Many psychologists already appreciate the above, I know. I’ve never seen a formal argument using Aristotle’s framework, but the broad point isn’t new. All I’m advocating is that we consider anew what we really mean by evidence-based practice, and perhaps those advocates and dissenters who have sometimes gone to war with each other in pages of our learned journals might come to a speak in a common language.

Leave a Reply

Your email address will not be published. Required fields are marked *