September 28, 2009
In a word, no. Quite the contrary, as we shall see below.
This post was inspired by apgaylard‘s epic fisking of a paper by Lionel Milgrom (Milgrom 2009), where Milgrom talks about the scientific basis of homoeopathy. While perusing Part 3, one of the papers apgaylard discussed caught my eye: van Wijk R, Bosman S, van Wijk EP. (2006) Thermoluminescence in ultra-high dilution research. apgaylard judges it to be one of the „better“ papers; it contains a statistical analysis of experimental results. After I left a comment on his blog, apgaylard very kindly sent me the paper itself. I set to work, looking at the statistical analysis and reanalysing the complete data set, which the authors kindly provide in their Table 1. What follows are comments from a statistical point of view – they are independent of the method of thermoluminescence itself. I am not a statistician, but have an interest in statistical analysis and experimental design.
September 27, 2009
I hope to post more again soon – ideally two or three times per week. I’m about to start two new exciting projects, am preparing my first submission to an open access venue, and working on a major revision of another two.