NWO Open Competition for replication research: 'Deals with fundamental question in my field'
Innovation through repetition. That is how university lecturer Jurriaan Witteman describes his research on the automatic processing of angry voices in the brain. The original research was conducted 20 years ago, but, with an NWO grant, Witteman is now going to see if those results are accurate.
The research focuses on a small area in the brain: the amygdala, also known as the almond nucleus. The current theory is that the area is a kind of threat detector. 'It could be a threat if someone talks in an angry tone: an angry person is also potentially dangerous. According to science, the amygdala is important for detecting that kind of threatening speech,' Witteman explains.
Unreliable results
The theory is popular within neurolinguistics, but there is one problem: it is questionable whether the theory is correct. 'The literature shows that studies tend to confirm it, but there is a tendency to confirm existing ideas, so results that agree with the hypothesis seem to be published rather than those that do not support it.' This can mean that for every published study that supports the hypothesis, there may be 10 contradictory unpublished studies that find no evidence for the theory.
In addition, studies within the field often use samples that are too small. For example, the original study that Witteman will now examine had a sample size of only 15 people. 'That is too few to detect effects, which could make the results unreliable.' At the same time, contemporary theories are based on the most prominent study. Witteman believes this is a bad thing: 'It addresses a very fundamental question in my field: are the results in neurolinguistics replicable and therefore do they really exist?'
Pure coincidence or not?
If the results are not fully replicable, it means that the theories that built on the original research may need to be modified. To find out, Witteman will use a larger sample to see to what extent the original results are replicable. 'One study can always be a coincidence. You only have an indication that the effect really exists when you find a result repeatedly. That’s why it is important to replicate authoritative studies,' he says.
The replication study is yet to be conducted, but Witteman has mixed feelings about the results. 'I hope that the results of the replication study agree significantly with the original results, but given the small sample size, I’m afraid that not all the effects will be replicable.'