We consider others more pleasant and trust them more when their emotional expressions confirm our expectations. This is the conclusion reached in a study conducted by researchers from the UAB and Northeastern University, Boston, who studied whether our predictions influence how we value others. The study was published recently in Journal of Personality and Social Psychology.
The study is based on emerging theories on the role predictions have in perception and action, especially in neuroscience. These theories indicate that our brains function in a predictive manner, with an internal model being constantly updated according to experience gained, without waiting to obtain information through the senses to interpret it. This seems to help process information more efficiently, but also conditions the way it is incorporated into our internal model.
Researchers conducted the study with the aim of verifying whether this predictive processing has a real effect on how we understand and experience the world. The research focused on a highly relevant field as is the role of emotions in social perceptions, which can be applied to other disciplines such as studies on multiculturalism and mental disorders.
The research was made up of six experiments with a total of 270 participants who were tested for their expectations on faces expressing happiness, sadness and fear in changing real-time situations and under specific contexts.
Participants rated the individuals as more pleasant and trusting when their expressions coincided with their predictions. “It is known that smiling faces are rated more positively than others, but here we see that this is not so when the individual shows an expression which is not expected. We were even able to observe that when shown two stereotypical expressions of the same emotion, participants preferred the one closest to their own mental pattern", explains Lorena Chanes, Serra Húnter lecturer from the UAB Department of Clinical and Health Psychology.
Hillary Clinton and Donald Trump confirm the hypothesis
Researchers confirmed their discoveries in the real world with the two latest Unites States presidential candidates. There were 90 participants who, in addition to the other experiments, were asked a series of questions on Hillary Clinton and Donald Trump. In addition to how pleasant and trustworthy they considered each candidate to be, participants also responded the question of how predictable they seemed.
“The two candidates violate social stereotypes and that could make them be judged as less predictable. For example, Clinton for gender and Trump for decorum. The results confirmed what had been seen in the other experiments: the more predictable a candidate was considered to be, the more trust and sympathy he or she aroused", Lorena Chanes points out.
The study also served to show that when our predictions are consistent with the information we expect we are quicker to respond. This result is in line with the theories of predictive processing, which proposes that our predictions have an optimisation and efficiency role in the processing of information.
Emotional predictions and their effects on behaviour have been studied indirectly before, through phenomena such as stereotypes and emotional coherence experiments (between verbal and non verbal signs, for example), but they had not been discussed in terms of predictive processing. ”Stereotypes are related to connotations shared by social groups. Our study was conducted with facial expressions which in theory are not associated with any judgement. Nevertheless, the social cost of violating expectations seems to be high. Every day we make emotional predictions which we may not be aware of and which influence our social perception. We finally discovered that we are more conditioned by our internal model than previously thought".
Researchers are now working on the design of experiments to be conducted with the clinical population, to discover how people with mental disorders perceive and evaluate others. “There are theories suggesting that people with depression use their internal predictive model to evaluate aspects or other people, but with very little updating. That makes their model become reinforced, but at the same time they incorporate less information from the outside, when it is the balance between learning about what is new and anticipating the future what makes us more functional. Our study may help to confirm this hypothesis”, she concludes.