How do brain processes bias our decision-making?

Published: September 4, 2024

Our lives are defined by the decisions that we make. Coffee or tea? English or Mathematics? University or an apprenticeship? We spend hours deliberating over some decisions, weighing up the options and worrying about the outcome. Others, we make in a heartbeat by following our noses and trusting our guts. Professor Ralf Haefner from the University of Rochester, USA, is studying how our expectations influence how we interpret our observations of the world around us, and how this biases our decision-making.

Talk like a cognitive scientist

Ambiguity — the quality of being interpretable in multiple ways

Confirmation bias — the tendency to seek out and interpret information in a way that supports our existing beliefs, while ignoring or dismissing information that goes against our beliefs

Confounding variables — factors unrelated to an experiment that can have an unwanted effect

Perception — the process by which our brains collect and make sense of the information gathered by our sensory organs

Primacy bias — being more influenced by information received first (e.g., trusting your first impressions)

Recency bias — being more influenced by information received last (e.g., forgetting earlier information)

Temporal bias — A bias that influences our beliefs and decision making based on the order in which we receive information. Primacy and recency bias are both examples of temporal bias

What decisions have you made today? From the moment we wake up, to the moment we fall asleep, we are constantly making decisions. While some seem insignificant, others can be life changing. 

“Decision-making is one of the most important functions of our brain,” says Professor Ralf Haefner from the University of Rochester. We base decisions on information that we collect from the outside world, but the way we interpret this information can vary depending on our personal beliefs. Ralf is interested in how these beliefs influence our perception and decision-making. 

What factors influence our decision making? 

“The field of psychology has characterised many ways in which human decision-making is biased,” says Ralf. “My research is focused on temporal biases, where information is treated differently depending on when it was received.” 

For example, imagine that you are trying to decide whether to watch a film based on your friends’ opinions of it. The first friend you ask says they hated it, but the second friend says they loved it. You might be biased towards the first opinion you hear and decide the film is not worth seeing. This is primacy bias. However, you might be most influenced by the review you received last, and decide you do want to watch the film. This is recency bias. 

Primacy bias 

Primacy bias is an example of a broader type of bias called confirmation bias. “We tend to interpret new information as being consistent with our existing beliefs, even when it is not,” says Ralf. “This makes sense; when interpreting what someone is saying, it can be useful to consider what you have learnt from previous interactions.” 

Imagine that someone describes a film using an ambiguous word like ‘unique’. To interpret whether this ambiguous word is meant in a positive or negative way, it makes sense to consider whether earlier descriptions were predominantly positive or negative. If previous descriptions have been positive, then we are likely to interpret ambiguous words in a positive way. However, if previous descriptions have been negative, then we are likely to interpret ambiguous words as also being negative. 

In other words, first impressions matter. Whether positive or negative, they can influence how we interpret ambiguous information. 

However, the strength of the influence will vary depending on which words are used. The less ambiguous the words, the weaker the influence. Imagine that multiple people describe the film using both positive and negative words like ‘amazing’, ‘wonderful’, ‘rubbish’ and ‘terrible’. Since it is clear what each word means, the influence of our expectations is weaker. Yet, there is still ambiguity about whether the film is worth watching, despite the fact that none of these words are inherently ambiguous. Here, the ambiguity stems from the fact that the opinions contradict each other. 

While the primacy effect had been known for a long time, before Ralf’s work, no theory could predict how strong it would be in a given situation or experiment. 

How is Ralf studying these biases? 

Although the above examples use language to explain the different biases, Ralf’s research focuses on perceptual decision-making, where decisions are not based on opinions, but on direct perceptual information, such as images on a screen or sounds from a pair of headphones. He studies this in the lab with various experiments that assess participants’ abilities to interpret and seek new information. 

Just as words like ‘unique’ can be ambiguous, Ralf designed some experimental conditions to be ambiguous too. In these conditions, the images are blurry and the sounds are quiet. Other conditions are designed to mimic clear but contradictory information (i.e., ‘amazing, wonderful, rubbish and terrible’). In these conditions, the images and sounds are clear, but they are rapidly swapped and shuffled, so the overall impression is ambiguous. 

“By studying these biases in the lab, we are able to eliminate confounding variables, repeat experiments many times, and study the sources and nature of biases precisely,” says Ralf. 

What do Ralf’s experiments show us? 

“When our observed pieces of information are clear, but contradictory, we are less likely to be biased than when the observations are weak or ambiguous, but consistent,” explains Ralf. When participants in Ralf’s experiments were presented with ambiguous perceptual information, they were more subject to primacy bias. However, when the perceptual information was clear – even if it was contradictory – participants were more likely to show recency bias, or no bias at all. 

“Our research also shows that in situations when confirmation biases, such as primacy bias, are strong, we’re more likely to be overconfident in our decisions,” explains Ralf. In other words, when the perceptual information is ambiguous, we are more likely to favour information that supports our own beliefs and to be overconfident in the decisions we make. 

Why is it important to study how humans make decisions? 

Reference
https://doi.org/10.33424/FUTURUM524

Dr Ankani Chattoraj has worked with Ralf to co-author multiple papers about bias and perceptual decision-making.
Dr Richard Lange co-authored a paper with Ralf and Ankani about confirmation bias in perceptual decision-making, as well as other papers on its neural basis.
Cognitive scientists often use machine learning techniques and artificial intelligence to mimic human thinking processes.
When the information we receive is ambiguous, we are more likely to succumb to biases.
Confirmation bias is a tendency to seek out and favour information that is consistent with our existing beliefs.
Ralf and his team at the annual Vision Sciences conference in 2022.

“The main reason we make decisions is to decide between different actions,” says Ralf. “In fact, the only reason we have a sensory system and the ability to perceive things is ultimately to make better decisions about what to do next.” Studying how this decision-making process works is vital to understanding more about the human brain. 

“Understanding how individual humans make decisions, and what limits or enables us to make good decisions, is also important for understanding our collective decision-making,” explains Ralf. This is particularly true in a political context, when large groups of people come together to make important societal and economic decisions. “It would be interesting to study which conditions amplify or mitigate the confirmation bias in this context,” adds Ralf. Political decisions have the power to change the world, so understanding how they are affected by biases could have a huge impact. 

Professor Ralf Haefner
Department for Brain and Cognitive Science, University of Rochester, USA

Fields of research: Computational neuroscience, brain and cognitive science

Research project: Understanding the relationship between perception and confirmation bias

Funder:  US National Institutes of Health (NIH), National Science Foundation (NSF)

About cognitive science

Cognitive science explores the complexities of the human mind and examines how we perceive the world around us, exploring topics including memory, decision-making, attention, reasoning and language. 

“Cognitive science is fundamentally interdisciplinary, providing many interesting research opportunities at the interface of behavioural economics, political science, psychology, neuroscience and artificial intelligence,” says Ralf. This combination of subjects and the collaboration with other colleagues can make cognitive science an exciting field to work in, full of complexity, challenge and stimulation. It also provides opportunities to learn about a wide range of subjects and experiment with areas that you may want to specialise in as your career progresses. 

A lot of cognitive science work takes place in laboratories with human participants and can involve techniques such as monitoring brain activity or eye movements as they carry out tasks or answer questions. Cognitive science can often involve creating computational models and using machine learning to enable artificial intelligence to mimic human thinking processes. 

For Ralf, working in cognitive science provides a number of rewards. “The experience of gaining new insights and making progress towards answering interesting questions, developing models or designing new experiments, and working with my colleagues, students and postdocs are the best parts of my job,” enthuses Ralf. 

Cognitive science does come with challenges, however. Research in cognitive science often involves designing experiments using human, and sometimes animal, participants. This can present ethical issues including thinking about privacy and consent, and how to conduct the research in a responsible, respectful and considerate manner. 

Keeping up with advancements in the use of technologies such as artificial intelligence, neuroimaging techniques and programming languages is exciting and an excellent launching pad for careers outside of academia. 

Pathway from school to cognitive science

At school, study subjects related to cognitive science including biology, mathematics, computer science and psychology. Getting familiar with coding will give you confidence when it comes to programming experiments, and learning about statistics and data will help with analysing results. 

Exploring other subjects including neuroscience, artificial intelligence, linguistics, philosophy and anthropology will deepen your understanding of the human mind and how it works. “Look for fun projects to do,” says Ralf.

“Look for likeminded people, and for competitions or science fairs to present your work.” Think of experiments you can carry out to learn more about the human mind and explore topics like human memory, decision making, attention and perception. 

Explore careers in cognitive science

Learn more about the research being carried out by Ralf and his colleagues at the Department for Brain and Cognitive Science at the University of Rochester.

Watch the videos and read the articles on the Cognitive Computational Neuroscience website . Ralf especially recommends the ‘GAC’s which are collaborations on controversial and important research questions.

Explore articles on the website of the Cognitive Science Society, covering the latest research being carried out in all areas of cognitive science. You could even consider becoming a student member.

Studying cognitive science can lead to a wide range of careers. You could become a researcher like Ralf, or you could follow a career path in fields such as robotics or advertising. The University of Rochester has provided a list of possible career opportunities related to cognitive science.

Meet Ralf

During high school, I wanted to become either a physicist or a computer scientist. I loved mathematics, physics and computer programming. 

The projects that I remember from my high school times include experimentally studying the conditions under which a rotating boiled egg stands on its tip and falls down again as it slows (involving lots of broken eggs), and simulating the movement of a planet around a binary star with beautiful orbital patterns. 

During my physics PhD, I changed tack and decided to do a master’s in politics and business to work in the ‘real world’ and have ‘real impact’. 

After my master’s in politics and business, I worked for a management consulting firm but did not like the unreasonably stressful life, working intensely for 80 hours or more each week with little time for friends, family or hobbies. I also wasn’t fulfilled by never having time to think deeply about things. I was mostly rushing around to get things done. 

After entering neuroscience, I slowly fell in love with the topic of understanding the brain (something I had thought very little about before) and really appreciated the generally welcoming dynamic and social atmosphere. 

Within neuroscience, I started in an experimental neuroscience lab at the National Institute of Health (NIH) and subsequently gravitated to more and more theoretical questions, becoming a computational neuroscientist in the process. 

My hopes and aspirations for the future are gaining a better understanding of how the brain performs ‘probabilistic inference’ (how we form and update our beliefs), the computations underlying our sensory perceptions and perceptual decision-making. 

I like spending time with my family and friends. I also love music, being outdoors (particularly running) and have always loved to travel. I also enjoy photography and playing with my remote-control plane, drone and cars! 

Ralf’s top tips

1. Find areas of life that you feel passionate about and that you are good at, and look for fun projects to do.

2. When trying to decide between two fields or directions to pursue, go in the more quantitative direction.

Do you have a question for Ralf?
Write it in the comments box below and Ralf will get back to you. (Remember, researchers are very busy people, so you may have to wait a few days.)

0 Comments

Submit a Comment

Your email address will not be published. Required fields are marked *

 

 

Learn more about biases and how this knowledge can be used to develop machine learning algorithms:

www.futurumcareers.com/bias-assumptions-and-emotions-why-we-think-what-we-think