Can we use nerve cells to gauge confidence in decisions?

We are all constantly making decisions: from simple tasks such as deciding when to get up and what to eat, to more difficult dilemmas, such as whether to invest in the latest gadget or making career choices. What separates easier decisions from more difficult ones? Researchers at the University Hospital Bonn have found that particular nerve cells in the brain are associated with levels of confidence in our decisions. 

The study involved 12 participants who were asked to complete two tasks. In the first, they were shown ten salty and ten sweet items (such as crisps, biscuits, and chocolate) and asked to rate them. Each item was presented 3 times, for a total of 60 trials. For the next task, participants were shown all possible pairings of the same 20 items, adding up to a total of 190 pairs. They were asked to select their preferred snack and instructed to move the slider in the middle of the screen in correspondence with how confident they were in their decision, so that the closer they moved the slider towards the chosen snack, the more confident they were. 

In reality, our psychology is much too complex to be completely characterised by one approach

During the process, a total of 830 neurons in the medial temporal lobe were being monitored through microelectrodes implanted in various regions of the brain including the hippocampus and amygdala. There are three main aspects of a decision which can be studied: confidence, choice, and reaction time. Researchers found a correlation between decision confidence and the frequency of electrical pulses being fired in certain neurons. This was especially clear in the second task: as participants’ confidence changed between experimental sets, so did their neuronal activity. 

Measuring the activity of particular neurons can give us an indicator of decision confidence

However, many of us know that correlation does not equal causation, and indeed the researchers note that their methods did not take into account that other variables could have influenced the results but were not detected due to the statistical methods used. For example, a small group of neurons could have very strong correlations between their activity and a cognitive variable other than conscious reports of confidence, and this could go unnoticed. This is a consequence of choosing statistical testing methods which reduce the risk of false positive results, at the expense of not being able to detect more nuanced activity, since hundreds of neurons across a relatively large area were being monitored. What can still be concluded, however, is that measuring the activity of particular neurons can give us an indicator of decision confidence.

The research presents intriguing findings, yet there is much still to be confirmed

The method of monitoring brain activity (electrode implantation) used in this research is very invasive, so there tends to be ethical implications preventing studies of this nature. The participants involved in this study were already undergoing the procedure since they suffered severe epileptic seizures that originated from one area of the brain. Electrode implantation is used to identify this location before surgical removal which meant researchers could also carry out the trial. This explains why the number of participants involved in the experiment is limited to a small number. Studies with fewer participants, especially when participant diversity is limited (e.g. by the fact that all had severe epilepsy), tend to be less representative of the population and thus it is more difficult to infer results from the data.  

This research reignites the debate of the level of influence biological phenomena have over our behaviour

The research presents intriguing findings, yet there is much still to be confirmed. Further studies are required to explore this association and level of accuracy between the persistent firing rates of neuron subpopulations and decision confidence. Different statistical approaches, perhaps focussing on smaller groups of neurons, can help solidify our understanding of neural activity during decision-making. 

In the meantime, this research reignites the debate of the level of influence biological phenomena have over our behaviour. Do neuronal firing rates affect decision confidence as well as indicate it? In reality, our psychology is much too complex to be completely characterised by one approach – be it biological, environmental, or other – but even so, each perspective gives us a unique insight into the multitude of forces driving our actions.