A reader recently sent me an account of a case which demonstrates the placebo effect by proxy and the general unreliability of our subjective assessments of medical interventions. Her story is engaging and nicely illustrates these common issues, and she has kindly agreed to allow me to reproduce it here. Enjoy!
AN EXAMPLE OF HOW TO FOOL YOURSELF – AND HOW THE PLACEBO EFFECT CAN WORK IN ANIMALS
Once upon a time, a horse owner said to an alternative therapist: “Thanks very much for treating Billy last week. He was much more relaxed than usual in the stable that night, and he went really well when I rode him the next day – more forward going, more supple and more willing than usual. I’d like you to come and do him again”.
Very satisfactory for all concerned. The horse was going well, the owner was happy and the therapist had a new client. Except for one thing – the therapist hadn’t treated Billy last week. She’d gone to the yard as requested and met not the owner but the groom, who through a misunderstanding had asked her to treat another horse. The owner, not knowing this, had ridden Billy the next day and had attributed her good ride to the treatment she thought the horse had had the day before.
Now this alternative therapist had an enquiring and scientific mind and decided to conduct an experiment. She asked a friend of hers whether she could give the friend’s horse a free session of her therapy. She didn’t want the friend to watch what she did, but she did want the friend to give her feedback afterwards. And she didn’t do anything to the horse at all. While the friend thought she was doing the treatment, she was actually sitting in the manger reading a book and the horse was eating his hay. When the therapist later asked the friend what had happened, she was not altogether surprised to hear a tale about a very relaxed horse who “went so much better than usual when I rode him the next day”.
Unlike most stories that begin ‘once upon a time’, this one is true. It illustrates very nicely the danger of attributing a change in our horse’s behaviour or performance to something we have just done. Or, as in this case, that we think we have done.
The therapy in question, unlike many, actually had some biological plausibility: it was a manipulative technique that many people find relaxing and invigorating, and it is not beyond the bounds of possibility to suppose that horses might also get at least temporary benefits. The point is that the intervention (or non-intervention) could have been anything: a veterinary treatment; a new feed or feed supplement; a new saddle or bridle; a visit from the farrier or equine dentist – you name it. The only necessity was for the owner to believe that any improvement in the horse’s behaviour following the intervention must have been a result of that intervention. As we have seen, the intervention didn’t even have to happen. All that was required was for the owner to want to make that connection.
The fact is that a horse will vary from day to day in how lively, enthusiastic, supple or willing he feels and there can be many reasons for this: working hard the day before, a slip in the field, more time than usual in the stable, the weather, the time of day, his social relationships and the amount of sleep he’s had, to name only a few of the possibilities. It is just unfortunate that we, with our pattern-recognizing, all-too-human brains, are likely to come to the wrong conclusions about the reasons why, because of what we want to believe. If we have given the horse something we believe will improve his performance – and especially if we’ve paid a lot of money for it – we are likely to think that any subsequent improvement in performance is a result of what we did. As you can see, the fact that a change in behaviour followed an intervention does not prove that the intervention caused the change, and this is why the scientific method is necessary.