Presentation ‘Stop Testing Null Hypotheses’ at NIP-Symposium
The photo was taken just before my PhD-defense almost 7 years ago and I still fight against the (mis)use of p-values and null hypotheses. For example, At the symposium When culture matters:
When: 2nd of November 2017
Where: De Koepelzaal in Het Huis, Utrecht, Boorstraat 107
We do know quite a lot about cross-cultural research, isn’t it? Why do we still bother testing the null hypothesis which asserts that there is no difference in the population? Even if we would not expect any difference, which is by the way a good hypothesis, the p-value would not give us an answer at all on this question.
In this presentation, I will explain in which cases the null hypothesis and p-value, when used correctly, would give a correct answer. I will also show how often we abuse the p-value and I will give a reason why we should stop doing so. In addition, I will show how Bayesian statistics work and what type of solutions it could offer. I will demonstrate through examples from cross-cultural psychology how the respective questions could have been answered in a better way. I will also show that also Bayesian statistics can be abused.
This presentation forms part of the project “The Living Room of Science” from the Young Academy of the KNAW, where a discussion is being carried out about a responsible way of making science.
P-Values and Null Hypotheses?
Want to know what p-values really are? Check the videos on my site.