Overview
The relationship between medicine and philosophy has a distinguished history. Maimonides, Avicenna, Galen, Descartes, and Locke were all philosophers and medical doctors. More recently, Peter Medawar and Archie Cochrane were strongly influenced by Karl Popper. There is an increasing body of evidence that combining History and Philosophy of Science on the one hand, and health care on the other creates synergies for the mutual benefit of all disciplines.
The course will consider:
- How and why did the idea that comparative studies were necessary to inform health care decisions replace other ‘methods’ such as reasoning from more basic sciences and ‘expertise’?
- Can average results be applied to individuals?
- What is the role of values?
We believe that the history and philosophy of science is an integrated discipline, and we will explore these issues with appeal to current and historical examples.… it is fair to say that not very much attention was paid by the originators of EBM to the philosophy of science… One hopes that the attention of philosophers will be drawn to these questions (Haynes, 2002)
A wise man proportions his belief to the evidence – David Hume
History of science without philosophy of science is blind … philosophy of science without history of science is empty – Norwood Russell Hanson
Description
Why study the history and philosophy of Evidence-Based Health Care (EBHC)?
- Deepen your understanding of EBHC. Just like understanding a country’s history can help you understand its current culture, so understanding how the ideas associated with EBHC arose will help you understand EBHC. Moreover, historical anecdotes are often easier to grasp than numbers.
- It is intrinsically interesting. Many people simply like delving into the history and philosophy.
- Become a better health care professional. There is a growing body of evidence that medical students who study the humanities perform better than those who focus exclusively on the sciences (Lancet 1996 (347:55-6), J Med Humanities 2004(30:53)). Explaining diagnoses, prognoses, and treatment benefits clearly to patients is important. You will developcommunication skills by both informal and formal presentations of your work, and by writing an extended essay. You will also improve your ability to think critically. Delivering Evidence-Based Health care is rarely straightforward. Dealing with complex scientific issues, writing research papers, teaching and presenting at conferences are now integral parts of clinical practice
- Become a better historian and philosopher of science. A team of philosophers, historians, and medical practitioners delivers this course. Philosophers and historians will therefore have the unique opportunity to study the history and philosophy of a particular science – EBHC – from the ‘inside out’.
- Why choose this course if you are neither a health care practitioner nor a philosopher nor a historian? Lay press bombards us with claims that diets will cure disease, that new medical technology promises to prolong life, and that environmental factors will kill you. Because of the interdisciplinary nature of the course, technical jargon particular to each discipline will be entirely avoided. Hence the course is accessible to anyone interested in understanding the nature, history, and justification for the kinds of claims they face on a daily basis.
You can find further details, costs, etc. at: http://www.conted.ox.ac.uk/courses/details.php?id=B900-77
Error Statistics Philosophy 
Hi, Mayo and Spanos.
Let me ask you something not related to the topic… have you heard of Alexandre Patriota’s work?
Click to access 1201.0400.pdf
He tries to define a frequentist measure of evidence.
Carlos: True, not on topic. But quick reply: Never heard of him, and on a fast scan, I can see this is just the nth rehearsal of the same old arguments, leading to accounts that fail utterly to satisfy severity requirements. See, for a short example:
Mayo, 2003 “Could Fisher, Jeffreys and Neyman Have Agreed? Commentary on J. Berger’s Fisher Address, Statistical Science”, 18, 2003: 19-24, 2003.
Beyond that, here are a few more that deal with the same issue (of p-values conflicting with posteriors).
Mayo, D. (2010). “Sins of the Epistemic Probabilist: Exchanges with Peter Achinstein” in Error and Inference: Recent Exchanges on Experimental Reasoning, Reliability and the Objectivity and Rationality of Science (D. Mayo and A. Spanos eds.), Cambridge: Cambridge University Press: 189-201.
Mayo and Spanos (2011): 187-9 (on blogsite).
All of my papers can be found from my site.
I will revisit this soon as the next joke of “the comedy hour” at the Bayesian retreat.
Ok. I have just bought your book (kindle version), I’ll check it out. The other references I have on pdf too!
Very good!
Just to let you know that the paper: “A classical measure of evidence for general null hypotheses”, cited by Carlos, is now accepted for publication in Fuzzy sets and systems. The final version was uploaded on arxiv: http://arxiv.org/abs/1201.0400
An interesting post from the Eran’s blog relating one of the “problems” of the p-values that I described in this paper:
http://eranraviv.com/blog/on-p-value/
It was re-posted at the R-bloggers:
http://www.r-bloggers.com/on-p-value/
Best regards,
Alexandre.
Alexandre: Yes, that is why error probabilities of tests, and hypotheses within tests, do not follow probability logic. For a blatant example, both H and not-H can be very poorly corroborated. With error probabilities it’s a matter of how easy (probable) it is to be wrong, and a test that has many more ways to go wrong, or many more ways to get results by chance, be it from asking more questions, or selection effects or the like, has altered error probabilities. The is the wonderful thing about error statistics, and not at all something to wish different. It is not an evidential-relation account, it completely throws overboard the “logicist” assumptions of logical empiricist accounts of evidence. That may not have trickled down to the formal philosophers as of yet.
Mayo, thanks for your response:
In this paper, I am not against the use of p-values, actually I just redefine them in a formal fashion in order to understand what is happening. The thing basically is: a statistic that has optimal behavior (in some sense) should vary with the null hypotheses (as you know). That is to say: it has optimal behavior for THAT specific null hypothesis. I know that you are aware of this fact, but when we write:
P(T > t ; under H0),
We are feeding many controversies, since it is not well defined (see the introduction of the paper: http://arxiv.org/abs/1201.0400).
A new measure of evidence, that *complements* the p-value and severity measures, with interesting interpretations on grounds of confidence regions is proposed and studied. It is important to say that p-values are not demonized in the paper, however if one also wants a support measure over Theta that does not depend on prior probabilities, the proposed measure can be successfully used.
It should be said that ma-many practitioners (in medical fields) use statistical hypothesis testing in their studies and they use nested hypotheses a lot. Well, p-values cannot be used in such contexts and they deserve to know it. What do you think about this?
Alexandre: I think maybe you should move your comment to the current post–I mentioned it there–so that others can more readily be aware of it.
Mayo, thanks for your response:
In this paper, I am not against the use of p-values, actually I just redefine them in a formal fashion in order to understand what is happening. The thing basically is: a statistic that has optimal behavior (in some sense) should vary with the null hypotheses (as you know). That is to say: it has optimal behavior for THAT specific null hypothesis. I know that you are aware of this fact, but when we write:
P(T > t ; under H0),
we are feeding many controversies, since it is not well defined (see the introduction of the paper: http://arxiv.org/abs/1201.0400).
A new measure of evidence, that *complements* the p-value and severity measures, with interesting interpretations on grounds of confidence regions is proposed and studied. It is important to say that p-values are not demonized in the paper, however if one also wants a support measure over Theta that does not depend on prior probabilities, the proposed measure can be successfully used.
It should be said that ma-many practitioners (in medical fields) use statistical hypothesis testing in their studies and they use nested hypotheses a lot. Well, p-values cannot be used in such contexts and they deserve to know it. Don’t you think?