Barking Mad - bad science in action

Faith was defined by St Augustine thusly: faith is to believe what you do not see; the reward of faith is to see what you believe. In 2000, in volume 321 of the British Journal of Medicine, appeared two articles that appeared to be directly contradictory of each other. Both were examining the hypothesis that dogs bite more often during the period around the full moon, and both used meta-analysis of hard data to reach their conclusions. The first study was conducted using admissions data from a hospital in Bradford, UK, while the second used data from several Australian hospitals. I am not proud, but also not surprised, to say that the dishonesty (for it can be nothing else) comes from the Australian team, or rather “Dodgy Duo”, of Chapman and Morrell.

The Bradford data examined all animal bites, and found conclusively that the period of the full moon consistently related to peak occurrence of animal bites, and that the peaks were extremely pronounced and unmistakable. The evidence clearly showed that there was a regular and pronounced increase in animal bites of all kinds as based on admissions to a hospital emergency room.

Now here comes the bad science; the intentionally misleading science; the misdirection away from evidence that does not conform to the desired 21st century ‘scientific’ paradigm. As can be seen from the title of the paper – Barking mad? Another lunatic hypothesis bites the dust – the paper was scathing about the notion that animal bite occurrence increased at the time of the full moon, and they had the data, and the graphs to prove it. And indeed they did. Their data and graphs clearly showed that the full moon was the obvious low point for animal bites in Australian hospitals, and that should have been that until the data, and the graphs, were looked at in detail. What becomes immediately obvious upon comparing the Australian graphs to the Bradford graphs is that they are virtually identical; except for the fact the high bite point is the full moon in England, and the new moon in Australia. Conversely, the low bite point in England is the new moon, and the full moon in Australia. Otherwise, if the two charts were overlaid, they are identical (or near enough to).

The obvious conclusion here is that lunar cycles are robustly shown to affect the incidence of animal bites as evidenced by hospital admission data, but that in the northern hemisphere, peak appearance is the full moon period, and in the southern hemisphere, peak occurrence is the new moon period. This in itself is a significant observation with myriad implications that deserve further research – hardly worthy of the grossly erroneous and intentionally dishonest epithet “another lunatic hypothesis bites the dust”.

Bhattacharjee C et. al (2000) Do animals bite more during a full moon? Retrospective observational analysis. British Journal of Medicine 321:1559-61.

Chapman S, Morrell S (2000) Barking mad? Another lunatic hypothesis bites the dust. British Journal of Medicine 321:1561-63.

Silverstone M (2011) Blinded by Science. Lloyd's World Publishing, London.