The intersection of bias and technology

Fay Cobb Payton has an extensive backround in IT. Returning to her alma mater, Payton spoke on bias in technology last week. // Photo courtesy of North Caroline State University

On March 10, Fay Cobb Payton, Professor Emeritus in the field of information technology (IT) at North Carolina State University, spoke as a guest of the School of Public Policy.

Her lecture centered around data bias and the importance of implementing an interdisciplinary approach to the collection and analysis of data in her presentation “Coding, Coded, & Counting: A Bias Continuum.”

Payton has worked in the field of data and information analytics for much of her life, with a particular focus in the healthcare sector. 

She attended the Institute for her undergraduate degree, earning a bachelor’s in Industrial and Systems Engineering.

During her time at the Institute she was a member of the National Society of Black Engineers (NSBE) and a participant in the Atlanta University Center Dual Degree Engineering Program.

When discussing her experience at Tech, Payton recalled her convocation speech which called on students to look in every direction and subsequently informed them that in four years only one of them would be left. The speech had a lasting effect on Payton.

She explained that this made her feel unwelcome in the competitive environment, and this was a very formative experience that has deeply impacted her approach towards the teams she works with.

She earned a bachelor’s in Accounting with a minor in Mathematics from Clark Atlanta University as part of the Atlanta University Dual Degree Engineering Program. 

Additionally, she earned her Master of Business Administration (MBA) from Clark Atlanta University. From there, she went on to then earn her Ph.D. from Case Western Reserve University in Information and Decision Systems. 

She currently serves as a Professor Emeritus at NC State University in Information Technology and Analytics, where she has been named a University Faculty Scholar for her extensive research work.

In her presentation “Coding, Coded, & Counting: A Bias Continuum,” Payton discussed how crucial the integration of socially determining factors is into medical care, and how she has spent her time in data research to support this idea. Payton recounted the ways in which her life experiences outside of academia have shaped her passion for equity in healthcare. Growing up in Augusta, Georgia she attended a high school with a medical program. 

She first participated in an OB-GYN rotation in which she learned that being a healthcare practitioner was not her calling after watching a live birth, but she soon discovered a passion for data analysis after her supervisor offered it as an alternative.

As a young professional, Payton worked for a project that focused on a home healthcare system for the caregivers of patients with Alzheimer’s Disease. She emphasized the importance of the lessons that she learned from this experience, especially her takeaways about medical data in particular.

Payton emphasized the difference in self-reported data and data collected via other means, and how the disparities between the two sets can reflect larger findings than the data itself might. 

In the process of acquiring data from various healthcare providers and vendors, Payton learned that the process of sharing medical data is often the most difficult part of doing data analysis in healthcare.

Different providers and vendors define medical terms differently, categorize medical conditions differently, report patient charts and outcomes differently, etc. This variation leads to the process of combining data from different sources to be a significant undertaking.

Payton has been part of teams that have performed research into the nature of health disparities that tend to appear together and into other contributing factors.

In the process of determining comorbidities of type 2 diabetes, her team noticed ways that medical data can be misleading.

One of the factors they searched for when categorizing health outcomes was the length of hospital stays and noticed two trends distorting their data.

Hospitals were releasing patients they knew would need to be readmitted very soon to lower their length of stay and that when patients were transferred to separate institutions, their data collection ceased. 

Their conclusion was that women and people of color were the most likely to experience disparities in their outcomes with type 2 diabetes. She then participated in a meta-data analysis to observe the intersection of mental illnesses and HIV. 

Through this research, Payton noticed that the ways that different institutions report codes, especially codes that denote different mental illnesses, varies wildly. 

The difficulty with determining the true length of stay for patients continued with this particular data analysis.

Recently, before the global COVID-19 lockdowns, Payton studied the mental health crisis on college campuses with the goal of identifying potential strategies to mitigating the crisis.

The largest need that her team identified was a need for culturally relevant services for students. Her team also identified that race-blind services (services that were not catered towards students’ differing needs based on their racial experiences) heightened structural inequities experienced by students.

In the growing world of artificial intelligence (AI), Payton noted the importance of diverse development teams, to combat racial biases that have already been noticed in AI technologies. She concluded by emphasizing the way that “big data” (quantitative) has failed thus far by perpetuating inequities and biases in data analysis and emphasizing the importance of “small data” (qualitative) in the process of improving equity in healthcare and beyond.

For all interested, more information on Payton and her work can be found at