Reality or coincidence? We need to move from probability to evidence.
In studies, it is an old acquaintance, the p-value. But mathematician Peter Grünwald wants to get rid of it. The p-value is too susceptible to fraud and can lead to a distorted picture of reality and chance. That is why he wants to work with the e-value. De Volkskrant spoke to him about it.
'As a mathematician, p actually hurts my eyes', Grünwald says in the article. People have been working with the p-value for decades, but by now p stands for problems rather than probability. The article in the Volkskrant gives several examples of this: from prematurely discontinued research to p-hacking.
The e, which stands for evidence, is already used in, for example, the technology world (Netflix). But it is mainly p that is established. However, Grünwald sees opportunities to further develop and apply the use of the e-value. Because, he says in the Volkskrant, the p-value is really only suitable for simple surveys.
Want to know more?
Read the full article in de Volkskrant (€, Dutch only).
About Peter Grünwald
Peter Grünwald, professor of Statistical Learning at the Mathematical Institute (MI) and senior researcher in the Machine Learning group at Centrum Wiskunde & Informatica (CWI) in Amsterdam, has received a 2.5 million ERC Advanced Grant to further develop his research. Read more.
Banner: CWI