The scientists wanted to know the likelihood the abbess was correct in her claims, and was this level of accuracy just purely from chance? Here is how they explained their analysis:Share
Our statistical approach is based on the model of the game Battleship where a two-dimensional grid (similar to an Excel spreadsheet) is formed by all the claims known to the author on one axis and the herbs (those still used today) on the other axis. Modern herbal indications (medical uses) are represented as ‘ships’ which the medieval author tries to hit by randomly tossing a ‘missile’ into the grid. The hypergeometric distribution gives the probability that x ‘correct’ indications (‘hits’) could be drawn from the set of N herb/claim combinations with n ‘shots’, and the number of ‘ships’ (today’s herb/claim attributions) is M.They focused their study on 85 plants that are being used today for medical purposes. It found that there were 212 health claims by Hildegard from this group, and 30 of them would be correct according to contemporary standards. If she had been making the claims up randomly, only between 6 and 7 of her cures would have accurate.
The study finds the probability of this happening just by chance is 1 in 10,000,000. They conclude:
The hypothesis that Hildegard could have achieved her ‘correct’ claims by chance is to be clearly rejected on the basis of the highly significant level of our new statistical procedure. The finding from this approach that medieval medical claims are significantly correlated with modern herbal indications supports the importance of traditional medicinal systems as an empirical source.They add that European researchers should also be more open to the possibility that herbs might be responsible for a larger variety of remedies – typically these plants are now only associated with one or a small number of medical treatments. (Read more.)