E.g. leeches. ASK: Other strange practices?
Why do people pursue ineffective health practices?
People's desire to get well can lead them to believe in unfounded drugs or treatments.
But: emotions do affect recovery, and emotions can lead people to do what's needed
"After this therefore because of this"
B follows A, so A causes B.
E.g. I have cold, I take vitamin C, I feel better.
- placebo effect
- natural cure
E.g. draft -> cold
Need to consider all cells:
| cure | no cure | |
| treatment | ||
| no treatment |
Need to consider alternative causes, e.g. placebo effect
Vagueness: someone will get better
Cover failures by blaming the victim: people responsible for their own illness.
Law of similars: like cures like
From: http://www.skeptic.com/03.1.jarvis-homeo.html
Hahnemann's invention of homeopathy is reported to have originated with an experience in which he ingested a substantial dose of cinchona bark (the source of quinine) used to treat malaria. He noted that the symptoms he experienced were similar to those of malaria. He reasoned that since the remedy produced symptoms in overdose similar to the condition it was used to treat, this principle, his Law of Similia, could be used to discern the value of various medicines. He called this process proving a medicine. Promoters often misrepresent homeopathy as treating the "causes" rather than merely the "symptoms" of disease, but its reliance on the "proving" of remedies shows that homeopathy itself relies solely upon a symptom treatment.
Hahnemann's Law of Similia utilized the primitive view of monism that "nature is a unitary, organic whole with no independent parts" (Webster's) with inherent principles that like is like, like makes like, and like cures like. Monism is the basis of many ancient practices (e.g., eating the heart of a lion for courage), and holds that if one object resembles another they are alike in essence (like is like); idolatry in which carving a likeness of a god actually produces the god (like makes like); and folk medicine practices such as snakeroot being good for snakebite, because of their resemblance (like cures like). Hahnemann revived Paracelsus' Doctrine of Signatures, which declared that herbs would cure conditions or anatomical parts they resembled (Garrison, 1929, p. 206). The homeopathic Law of Similia, however, is unsupported by the basic sciences of physiology, pharmacology and pathology.
Ininitesimals: small doses; at least no one hurt.
Don't eat fruit with anything -> Fit for Life diet
Psychological problems caused by repressed memories of sexual abuse.
Memories implanted.
Chiropractors: disease caused by spinal misalignment
Note: there are good causal theories, but these have evidence,
e.g. low fat diet -> low heart disease.
Mind-body connections: stress & illness.
Good for prevention and looking after yourself.
But lots of dubious, untested ideas.
Acupuncture may be effective for pain and nausea, for reasons independent of the theory of traditional Chinese medicine.
Transcendental meditation lowers blood pressure.
But well-conducted scientific experiments are needed to determine whether a medical treatment or health practice is really beneficial.
Post hoc ergo propter hoc (Gilovich, ch. 8)
Tendency to infer that two events are causally related just because one happened after the other.
Kinds of claim:
A causes B. E.g. HIV virus causes AIDS. Toxins cause fatigue.
A cures B. E.g. antibiotics cure strep throat. Magnets cure arthritis.
A prevents B. E.g. exercise prevents heart disease. Garlic prevents colds.
Are they authorities?
What is their motivation?
None?
Anecdotes?
Tradition?
Case studies?
Controlled experiments? Evidence-based medicine (see below)
Consider all the relevant evidence of good quality, e.g. replicated (repeated) controlled experiments.
Does the claim (e.g. about a cure) explain the evidence?
What alternative explanations are there? E.g. placebo effects, incompetence, fraud.
Is the claim a better explanation than the alternatives?
Controlled: compare group getting treatment with another non-treatment group.
Randomized: assign patients to groups randomly so that there is no bias in who gets what treatment.
Blind: patients do not know what group they are in.
Double blind: doctors do not know what group patients are in.
Large: have sufficient group size that differences are statistically significant.
Objective: results are assessed and reported by responsible researchers, without fraud or misleading interpretations.
Doctor's Guide to the Internet
Updated Feb. 10, 2003