By Ingrid Fadelli, Medical Xpress
Four examples (four panels) of how different mental state inferences (y-axis) are shaped by first impressions of faces (x-axis) in samples from different regions of the world (indicated by different colors). Credit: Lin et al.
When we first meet another person, we typically form an initial impression of them based on their facial features and appearance. These first impressions of others can potentially influence our subsequent cognitive processes, such as which mental states we believe people we meet are experiencing at a given moment.
Researchers from the University of California San Diego (UCSD), the California Institute of Technology, and Dartmouth College conducted a study investigating the potential relationship between first impressions of faces and mental state inference. Their findings, published in Nature Human Behavior , suggest that first impressions of faces influence the inference of other people’s mental states.
“Over the years, there have been many surprising findings showing how first impressions of faces can predict important outcomes, such as which candidates will win an election, which politicians will be convicted of corruption, and which criminals will be sentenced to death,” Chujun Lin, the paper’s first author, told Medical Xpress.
“These findings show that the snap judgments people make about others based solely on their faces can influence important real-world decision-making, from who we vote for, to who authorities investigate, to how jurors evaluate cases.”
Because most people rarely engage in criminal investigations or legal trials, their initial impressions of others based on their appearance may not necessarily have a crucial impact on their decisions. Lin and his colleagues set out to determine whether first impressions also influence real-world scenarios that people engage in more regularly.
“We investigated how first impressions can shape the way people infer each other’s thoughts and feelings at every moment,” Lin said. “Understanding how each person feels and thinks is a crucial task in everyday life, as long as you interact with other people.”
A fundamental challenge associated with psychology research is conducting reproducible studies that produce similar results across different samples of participants, even in somewhat different settings. Lin and his colleagues therefore sought to devise reproducible and robust experimental methods that could be used by other researchers.
“To understand the complex relationships between face perception, mental state judgment, trait judgment, and situational effect, we used computational models to quantitatively select a large number of faces, mental state terms, trait terms, and situation descriptions that are representative of those people encounter in everyday life,” Lin explained.
“We asked participants to look at faces and infer how much these individuals would feel certain mental states in given situations.”
The researchers also asked a separate group of participants to look at the same images of faces and infer the traits of the people they belonged to. Using the information they collected, they then digitally manipulated the features of the faces.
Four types of mental state inferences (in each radar plot panel) are shaped by different types of first impressions of faces (vertices in each radar plot panel). Credit: Lin et al.
“We quantified the degree to which changing the perceived features of a face would change people’s expectations about how that individual might feel and think in different situations,” Lin said. “To ensure that our results can be applied to a wide range of populations, our data and models were based on participants from five continents: Africa, Asia, Europe, North America, and South America.”
Lin and his colleagues closely examined how study participants thought specific people in images would feel across 60 different mental states in 60 real-world scenarios.
For example, how much they thought a specific person would feel jealous if they heard that their best friend was admiring a new person, or how much they would feel lonely if they felt like a minority or different from everyone else in a group.
“We found that 47 of these 60 mental state inferences were shaped by the individual’s appearance (e.g., if you look feminine, people would expect you to feel more jealous if you hear [that] a best friend admires a new friend; if you look like a leader, people would expect you to feel less lonely if you are different from everyone else in a group),” Lin said.
“This means that in most circumstances, when other people try to understand how you feel and think, their understanding will be influenced by their first impressions of your personality (which is not necessarily your true personality, but just the judgments of others).”
Interestingly, the researchers found that first impressions shaped mental state inferences among participants living on all five continents on Earth. This suggests that their findings are robust and the effect they observed is relevant to all people, regardless of their nationality or cultural background.
“The goal of scientific research is to improve human lives, so it’s important that the way psychologists conduct research is relevant to real life,” Lin said.
“This may sound simple, but it’s not the case in our field. Most research has been conducted using highly controlled designs, such as having participants read vignettes and press buttons. But experimental designs that bear little resemblance to the real world are unlikely to reveal real-world psychological processes.”
Lin and his colleagues’ recent study could inform future research efforts focused on how first impressions influence what humans think about others and interpret their behavior or emotions. A key goal of his lab at UCSD, the IMPression in ACTion lab, is to continue gathering more naturalistic knowledge of how people understand themselves in the real world.
“We do this by promoting more naturalistic experimental designs (such as video watching and storytelling), leveraging more advanced computational tools that have been proven to work in the complex real world (such as large language models and vision models), and examining more diverse populations (participants from different regions of the world),” Lin added.
