1. Home
  2. Bias Bias Everywhere

Bias, Bias Everywhere

Article Date
October 07, 2022

Individual Bias and its Impact on the Prevalence of Scientific Bias

This article is a part of the DEI Resource Highlight series. 

Bias is defined as “prejudice in favor of or against one thing, person, or group compared with another, usually considered to be unfair.” One thing is certain, we all have biases. What happens when we’re not aware of those biases and we allow them to influence how we think and act?

The Southern California Clinical Translational Science Institute produced a resource titled Introduction to Clinical and Translational Research: Bias in Clinical/Translational Research: Analyzing Data and Interpreting the Results. Lectures cover potential sources of bias that can occur in all stages of resources and how to mitigate their impact on findings as well as analysis of data in biomedical research, its presentation, statistical testing, and conclusions.

Generally, there are two main types of bias—conscious and unconscious. Conscious bias is biased attitudes about a group that we are aware of. For example, in the case of affinity bias (a type of conscious bias) when we prefer people who share similar qualities to ourselves. Unconscious bias is prejudice or unsupported judgments in favor of or against one thing, person or group as compared to another. For example, attribution bias (a type of unconscious bias) that applies to how we assess others and their achievements.

Like an artist’s drawing mannequin, useful for mimicking different poses, if we train our brains, we can avoid perpetuating bias in our personal and professional lives.
Like an artist’s drawing mannequin, useful for mimicking different poses, if we train our brains, we can avoid perpetuating bias in our personal and professional lives.

Bias appears throughout the research process, too. Let’s focus on research publication where four primary research publication biases have been identified:

  1. Publication bias;
  2. Outcome reporting bias;
  3. Spin bias; and
  4. Citation bias.

Publication bias occurs when a study’s outcome influences the decision on whether or not to publish. Outcome reporting bias is illustrated when the author writes up only results in a trial that appear positive, while failing to report those that appear negative. Spin bias appears when the author uses language, often in the abstract or summary of the study, to make negative results or statistically insignificant results appear positive. Citation bias can occur after publication where a study is cited and discussed more, it is circulated more with positive results gaining more visibility and exposure than negative studies. All of these combined or even individually can taint the veracity of perceived or actual advancements in medicine in science.

Reflection

  1. Review The Cochrane Collaboration’s tool for assessing risk of bias in randomized trials HERE. Try applying this assessment to one of your current or past studies (if applicable).
  2. The most common biases are based on characteristics related to identity (e.g., race, ethnicity, gender, religion, sexual orientation, socioeconomic background, educational background). Think of a time you observed or prejudged someone based on one of the aforementioned characteristics and how it might’ve resulted in a discriminatory practice. How would you handle that situation now?
  3. How might you explain how individual biases impact scientific biases? What are some strategies you’d develop and implement to lessen the impact of those biases—individually and as a team?

Recommended Reading