Arggh, the girl I sit beside at work, who I thought was a fairly well adjusted, rational person disclosed to me the other day that they use hompathic remedies for their horses. Apparently they really work.
I pointed out how the placebo effect works on the humans as opposed to the animal and she said she knew that but they'd successfully treated physical conditions that nothing else would fix. I mentioned that there may have been an active ingredient in the homeopathic remedy which actually had an effect and she conceded that may be the case.
So of course we had a bit of a debate about homeopathy.
She already knew the 'science' behind it, but essentially her opinion was that things like this can work even though science can't prove it.
All of this is terribly familiar of course, but to me the interesting point was that I couldn't get her to understand the importance of blind trials. Or the idea that even though science can't explain how everything works, that using scientific principles to evaluate the effectiveness of treatments is the only way we can tell if they 'work' or not. And how it matters that we know whether conditions get better on their own with or without treatment.
I feel that maybe sometimes as skeptics constantly exposed to these ideas, we do not fully understand how difficult it is for the uninitiated to really grasp these subjects. Certaintly I've had this reinforced by this conversation.
I know how intellectually snobbish this is going to sound and I regret it, but ... I really do think we lose sight of how great the gulf is between we who can think and everybody else.
Yes, but it is true in many ways. Of course thinking that you are always right, spouting forth on subjects you don't really know that much about or generally being an arrogant tosser are traits to be avoided by anyone, knowledgable, intelligent or dumb.Originally Posted by DrS
I know how intellectually snobbish this is going to sound and I regret it, but ... I really do think we lose sight of how great the gulf is between we who can think and everybody else.
People are taught at an early age that magic is true and that wishing works, so is it hardly surprising that people have difficulty in understanding that no matter how much you wish or how much you believe in it will not make it come true.
Every one tells them that wishing and praying works and that make believe is real, how do you cure that. It's like alcies you got to want to be cured before you open your eyes to the truth that praying and wishing don't work.
I’ve got quite a few friends who are firm believers in homeopathy.
Whenever I try to persuade them that there’s nothing in it (literally) they come back at me with “It works on my pet cat, so it must be valid, because an animal can’t be biased”.
For years I didn’t have a convincing reposte to this statement, but now I think I have the answer.
It’s a statistical principle, called reversion to the mean.
Basically, the way that reversion to the mean works is that the further a thing is from its normal state the more likely it is to move back to it (as though it’s on the end of a piece of elastic).
So, the sicker an animal becomes the more likely it is to start getting better (or die).
People tend to get their animals treated (homeopathically or otherwise) when they are noticeably ill, so the animal is probably about to get better to some degree quite soon anyway (or to die).
That’s possibly all there is to it.
I came across this as a possible explanation while reading How We Know What isn’t So: Fallibility of Human Reason in Everyday Life - an excellent book by Thomas Gilovich.
There are other possibilities than regression to the mean: for example, while an animal doesn't know it has been given a treatment, neither can it report any subsequent improvement in its condition, which will be assessed by the owner and/or the person treating it, who will be well aware that it is "supposed" to get better - see for example this case report from the British Veterinary Voodoo Society's web site (incidentally, one of the best sources on the web for info about homoeopathy, especially in the context of veterinary medicine):And see this study: Rat models of acute inflammation: a randomized controlled study on the effects of homeopathic remedies, noting the "discrepancies" between the single-blind and double-blind phases of the experiment.The dog stood there, to my eyes actually slightly worse than it had been on the day I'd last seen it. Frankly, it looked just awful. But in the owner's eyes there had been a massive improvement.
Regression to the mean is only one form of bias here, as explained in this article:
The relationship between baseline value and its change: problems in categorization and the proposal of a new method.
Tu YK, Baelum V, Gilthorpe MS.
Eur J Oral Sci. 2005 Aug;113(4):279-88. Review.
PMID: 16048519 [PubMed - indexed for MEDLINE]
Also best not to forget an old favourite - confirmation bias.
http://www.ukskeptics.com/article.ph...nfirmation.php
Regression to the mean: http://www.ukskeptics.com/explanatio...ve_fallacy.php
.
When I worked with horses we knew of a woman who felt the best option for her difficult steed would be a horse whisperer (all fluffy and sweet sort of thing). Anyway the whisperer told her the horse said (should that be whispered?) it was distraught to leave his home and friends everyday when she was trying to ride him, the horse whisperer suggested it may take a while to solve the problem. So, anyway, she shot the horse because she couldn't bear the thought of it being so upset! That and she wasn't go to pay to feed a useless horse of course! On hearing the news we were somewhat stunned!
Poor bloody animals don't even get the privilege of being stupid in their own right!
Yes regression to the mean certainly works without bias and is more than likely in play in the example you give.
That's part of the effect.
Then there are all the other baises that are still present in animal experiments. Clinical trials are ideally double blinded. That is neither the experimenter not the patient knows wether the treatment is the test treatment or a placebo. In a single blind test the patient doesn't know but the experiemtner does. Often this may be necessary and not easily avoided - e.g. accupuncture. Sometimes it's because the experiment is just shoddy.
So what are these biases that a second layer of blinding can help with and are they applicable to your friend giving their cat a homeopathic remedy.
I leave the answer to that question as an exercise for the interested reader.
The keeper of our Irish Sport Horse has a problem with "conventional" vetinary practice and insists on prophylactic,that is,regular preventative,homeopathic "remedies". She is not covered by the insurance without MRCVS "MOT`s". Keeper won`t have it,what arguments to suggest? Horse is not ill or hurt yet,but that`s part of life as a mammal,dreading the day Jerry
Option 1) Get a different Keeper
Option 2) Engage in a long drawn out polite and well mannered discussion where you present the evidence against homeopathy both from the point of view of plausibility and also clincal trials. Through the slow moveing behemoth of logic, reason and impeccable etiquette convince the keeper that as it's your horse you get to dictate based upon reason and logic that only evidence based medicine is to be used on your horse.
Option 3) Lump it.
There may be other options. I doin't know much about the Gee Gees.
Option 4) Give the groom (say) 30 identical bottles (apart from a number whose significance is known only to you), one containing his favourite homeopathic medicine and the rest containing distilled water. If he can consistently work out which is the medicine, let him continue. If not, he either uses real medicines or looks for another job.
If they're "preventative", how can they fit them to the symptoms the horse doesn't yet have, and how can the remedies "encourage the body's healing process" by intensifying the symptoms? Use as prophylaxis flies in the face of what passes for theory in homoeopathy.
Perhaps there's a 'broad spectrum' homeopathic solution ...
Posting Permissions
Reply With Quote
Bookmarks