Title: Lessons from a life in applied statistics
Facilitator: Martin Bland, University of York, UK, martin.bland@york.ac.uk
Professor Bland is Professor Emeritus in the Department of Health Sciences at the University of York, UK. Personal research interests are in the design and analysis of studies of clinical measurement and of cluster randomised clinical trials. He is the author of An Introduction to Medical Statistics, now in its fourth edition, and co-author of Statistical Questions in Evidence-based Medicine, both Oxford University Press, more than 290 refereed journal articles reporting public health and clinical research and on research methods, and, with Professor Doug Altman, the Statistics Notes series in the British Medical Journal.
Workshop Description: Statistics can be applied to many things and the statistician may be called on to design the collection and analysis of many types of data. My main professional interests have been in agricultural trials and then epidemiological and clinical trials. These studies have been published in academic journals, but several other things have come my way which were unsuitable for this. I discovered that Significance, the magazine of the Royal Statistical Society rather than a journal, would publish things because they were interesting even if they were not conventional research. In this workshop, I shall describe several of these projects and look at some of the statistical lessons we can draw from these projects.
The Horizon homeopathic dilution experiment: Producers of a TV programme asked me to help analyse an experiment to demonstrate the “memory of water”, a repetition of a published experiment but with rigorous randomisation and blinding. We went on to break the blinding and for me to analyse the data live on camera.
Do the left-handed die young? Doug Altman and I were intrigued by studies purporting to show that left-handed people had shorter lives than the right-handed. We thought these studies were all methodologically flawed. I obtained a data set of 6,000 first class cricketers and carried out time-to-event analyses. The results were surprising.
Diabetes, blood sugar and red wine: a personal study: I used my own daily measurements of fasting blood glucose to test the casual observation that glucose appeared to be less on mornings after drinking red wine. Why find research participants when I am already here?
Risedronate, the BBC and Me: A radio journalist asked me to analyse a set of data from a trial of a bisphosphonate drug, obtained accidentally by one of the clinical collaborators. The analysis found very different results from those published. I looked for a possible explanation for the discrepancy.
Health freaks on trial: Duct tape, bull semen and the call of television: A TV company asked me to help to design and analyse a series of small clinical trials of obscure remedies promoted on the Internet. Four trials were done: oatmeal baths for psoriasis, oil-pulling, which is swilling the mouth with vegetable oil, for mouth problems, yoghurt and turmeric for acne, and duct tape for warts and verrucae. One remedy showed an effect, which was spectacular.
Promiscuous publication: Statistics can be applied to many different topics. In how many different journals can one statistician publish? A surprisingly large number, it turns out.