Extraverts exhibit greater error signals to reward prediction violations.

We all have those friends who are gregarious, confident and who seem to have an ability to attract conversation. Personality psychology would find that these tendencies (i.e. to be social and outgoing) all correlate and share an underlying relationship: that of belonging to the domain trait, extraversion.

Recent neuroscientific study has begun to tie the biological underpinnings of such extraverted related tendencies – like that of being bold, gregarious and captivating – to the neurotransmitter dopamine, which is often referred to as the ‘reward’ chemical. Recent research, however, suggests that dopamine isn’t just about encoding reward, it’s also about encoding the prediction of how much reward (if any) one should expect.

The seminal research that linked dopamine (at first) exclusively with reward, came from Olds and Milner’s research with rats. Their studies implanted electrodes into various parts of the rat brain. If the electrodes were to be stimulated, sending an electric pulse to the neurones that were situated near the implanted electrodes, the rat themselves would have to press a lever. In other words, the rat had complete autonomy regarding whether the electrodes would stimulate the brain. It transpired that when the electrodes were situated within the midbrain – where neurones primarily manufacture and communicate with dopamine – the rats opted to endlessly press the lever. The rats chose to press the lever, even to the detriment of consuming all other essentials, such as food and water. These rats would seemingly rather stimulate the release of dopamine than eat and survive. Hence, these dopamine producing neurones became known as the ‘reward centre’ of the brain.

Subsequent animal self-stimulation research finds that dopaminergic neurones, within the midbrain, primarily induce learning and approach behaviours. When these regions are stimulated, for instance, it becomes more likely that a rodent will approach an uncertain situation, when previously that same situation evidentially would have caused withdrawal or apprehension. The dopamine reward response, too, was found to hold in both monkeys and humans; when human participants are given money, a heightened midbrain dopaminergic response can be observed. As well, it appears these midbrain neurones are tuned to, not only the reward itself, but to the amount of reward received. Give a human participant a greater sum of money, and the level of dopamine activity will rise accordingly.

Closer study also revealed that the dopaminergic midbrain was partly responsible for our propensity to seemingly link a stimulus to a reward – AKA classical conditioning. For instance, studies have shown that, when a sound, image, or light are perceived before a reward is presented, the dopamine response occurs at the onset of the reward as well as after the onset of the conditioned stimulus (CS). Midbrain dopaminergic cells, therefore, encode both how great a reward is as well as what may predict the giving of a particular reward.

It appears, however, that the midbrain dopamine response is a little more subtle and intelligent than first thought.

Researchers began to realise that, proceeding classical conditioning, the presentation of a reward would no longer evoke a phasic dopaminergic spike in the midbrain. For instance, when a monkey has learnt a particular image regularly precedes a grape being given, the dopamine spike seems to be transferred to the conditioned stimulus (the image). That is, the dopamine response occurs after the stimulus (image) rather than after the giving of the reward (grape). Dopamine, then, can’t simply be just the ‘reward’ signalling chemical; instead, it appears midbrain dopamine cells are actually predicting the potential of a reward being received (Figure 1)

Figure 1: Midbrain dopaminergic neurones responding to a stimulus that is predictive of a reward.

Not only does it seem that dopamine is predicting a future reward, it also appears that dopamine is noting when a reward violates a prediction. Further research has shown that, when a CS is shown, but a predicted reward is subsequently not given, the dopamine response (at a time when the reward would usually occur) falls below a baseline (Figure 3). Whereas, when a CS is shown, and the subsequent reward is greater (Figure 2) than those given on previous trials (i.e. 2 grapes instead of 1), the midbrain dopamine cell response is heightened. Interestingly, The midbrain dopamine response appears to be predicting reward as well as monitoring when an error – relative to the prediction – occurs.

Figure 2: A midbrain dopaminergic positive reward prediction error.
Figure 3: A midbrain dopaminergic negative prediction error

These reward processing complexities are thought to partially underlie an individual’s propensity to exhibit extraverted related tendencies (or at least their greater propensity to demonstrate an ease toward approaching new situations). The reward-processing theory of extraversion purports that, a tendency to be bold, assertive and talkative is driven by an over-weighted prediction (in extraverts) that these situations will yield a reward. The midbrain dopaminergic system is, more precisely, predicting (in some way) that being talkative, and the centre of attention will make reward more attainable. Extrapolating further, this would mean that, those who score higher on measures of extraversion would be more likely to peruse social and adventurous endeavours – partially because of their more reactive dopamine driven reward predictions.

Smillie and colleagues (2019) – and prior studies – substantiated that there does indeed exist a relationship between reward predictions and trait extraversion (that is, a tendency to be talkative, assertive and more prone to positive affect). Smillie’s lab had participants passively observe an initial image (either a gold bar or lemon) that was 80 per cent predictive of a second image (again, either a gold bar or lemon). For example, if a participant is shown a lemon, there’s 80 per cent chance the next stimulus will also be a lemon and for the participant not to attain a reward. Conversely, if the participant is first shown a gold bar, then there’s an 80 per cent chance the next stimulus will also be a gold bar – indicating a monetary reward will be received. On the remaining 20 per cent of the trials (in both conditions), the first stimulus will not be congruent with the second stimulus (stimulus 1 = gold bar, stimulus 2 = lemon, non-reward given). These trials are intentionally set-up to elicit a reward prediction error signal from the participant (i.e. the midbrain dopamine prediction will be incorrect regarding the likelihood of receiving a reward/non-reward).

Figure 4: Smillie and colleagues (2019) associative learning paradigm. This is a predicted reward trial (S1 = gold bar; S2 = gold bar; accumulated reward).

Participant’s brain waves were measured from EEG channels located around the frontal scalp area. At around 200-300ms, after the participant has been shown an unpredicted reward/non-reward stimulus (i.e. S1: lemon — S2: gold bar — monetary reward), it is believed the midbrain led reward prediction error (RPE) signal can be measured as either a more negative (unpredicted non-reward) or a more positive (unpredicted reward) voltage deflection – relative to a baseline – at these frontal channel sites. This is exactly what Smillie’s lab found: An unexpected reward elicited a relatively less negative deflection wave, whereas an unexpected non-reward elicited a more pronounced negative deflection, which is indicative of a reduced midbrain dopamine response (Figure 5).

Figure 5: the average EEG wave deflection in each trial type.

Smillie and colleagues (2019) went onto investigate the relationship between extraversion and the reward prediction error associated EEG signal. Firstly, they calculated a reward positivity wave (the unexpected reward minus the unexpected non-reward wave) thought to encapsulate the dopamine related reward prediction error. The researchers found that extraversion was the only personality trait that significantly predicted the reward positivity wave; meaning, for every unit increase in an extraversion score, there’s a predicted increase in the reward positivity amplitude – indicating a greater RPE response to violations expectation.

These studies suggest that, individual differences in how extraverted an individual may be can be partially explained by the vigorousness of the midbrain-dopamine response to stimuli that are encoded as predictive of reward. It is thought that this more robust prediction of reward may underlie extraverts’ general tendency to more readily approach novel situations and seek social affirmation as well as popularity.  

Reference:

Smillie, L. D., Jach, H. K., Hughes, D. M., Wacker, J., Cooper, A. J., & Pickering, A. D. (2019). Extraversion and reward-processing: Consolidating evidence from an electroencephalographic index of reward-prediction-error. Biological psychology146, 107735.

Schultz, W. (2016). Dopamine reward prediction-error signalling: a two-component response. Nature Reviews Neuroscience17(3), 183.

For further information regarding this blog’s topic, contact the author, Thomas Cornish, at: tcorn001@gold.ac.uk.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s