You are here:
An Overview of Confirmation Bias in Science: Examples and Opportunities for Improvement
Citation:
Guiney, P., W. Goodfellow, AND Timothy J. Canfield. An Overview of Confirmation Bias in Science: Examples and Opportunities for Improvement. Society of Environmental Toxicology and Chemistry (SETAC) SCICON2 – SETAC North America 41st Annual Meeting, N/A, N/A, November 15 - 19, 2020.
Impact/Purpose:
Human thought processes are not perfect. We face cognitive errors daily. Confirmation bias is a common error committed by people, and contrary to the thoughts of many, scientists are not immune to committing this same error. It’s a tendency to believe that you are right, disregarding things that conflict with your ideas. Confirmation bias is reinforced with time. As scientists, we want the public to trust science, scientists and the data they put forth because they have confidence that scientists are intelligent, honest and follow procedures that ensures validity of the data and results. However, scientists are human and equally capable of succumbing to faulty thinking processes as everyone else. So, the onus is on the scientist and the scientific community to put forth the best and most unbiased data possible. The purpose of this presentation is to set the stage for the session about what confirmation bias is, how to recognize it, and approaches to minimize confirmation bias in the science that is presented, The tall will cover several examples of confirmation bias in scientific research and conclude with some ideas and recommendations on how to identify areas of confirmation bias, leading to opportunities to reduce or eliminate bias from our science. The expected impact of this talk is raising awareness of confirmation bias in science and providing ways to combat and eliminate it as much as possible in the presentation of science to the public.
Description:
Human thought processes are not perfect. We face cognitive errors daily. Confirmation bias is a common error committed by people, and contrary to the thoughts of many, scientists are not immune to committing this same error. It’s a tendency to believe that you are right, disregarding things that conflict with your ideas. Confirmation bias is reinforced with time. The longer you believe something, the more time there is to collect evidence in support of your opinion and the less willing you are to consider evidence counter to your opinion. Thus, it becomes increasingly more difficult to make real adjustments to the opinion or to abandon it completely. Confirmation bias also contributes to exaggerating belief probabilities. When you accumulate evidence in your mind that supports your hypothesis, you can believe something is much more probable than it actually is. As scientists, we want the public to trust science, scientists and the data they put forth because they have confidence that scientists are intelligent, honest and follow procedures that ensures validity of the data and results. However, scientists are human and equally capable of succumbing to faulty thinking processes as everyone else. For the public to be truly au fait about research conclusions, they should not simply accept the findings unless they have carefully checked the research and evaluated its quality and reliability. However, we know that most in the public are not equipped with the skill set or the desire to perform self-reflection and do this evaluation themselves. So, the onus is on the scientist and the scientific community to put forth the best and most unbiased data possible. Science is fundamentally about hypothesis testing where the scientist should have no personal preference for finding the hypothesis either supported or rejected. The pursuit to collect data that is without bias to test the hypothesis is the role of every scientist. Proffering hypotheses that one wishes was true but the data does not support causes the future pursuit of errant paths in science, a waste of limited resources in these deviant pursuits, and when it comes to light, erodes the support of the public for scientists and the science they produce. This presentation will cover several examples of confirmation bias in scientific research and conclude with some ideas and recommendations on how to identify areas of confirmation bias, leading to opportunities to reduce or eliminate bias from our science.