Coded Bias: The Impact of an Awareness Intervention on Coders and Computer Science Educators

• In Event: Developing Educators' Critical Literacy

Abstract

If an algorithm is applied to a dataset created with a racial, gender, socioeconomic, or other bias, its output will perpetuate the systemic bias shown in datasets used to train it. To ensure that the workforce is helping to develop algorithmic fairness, this study examines how undergraduate computer science, data science, and graduate technology education students react to an educational treatment around algorithmic bias. This treatment of a film and book discussion uses a Theory of Change. The pre- and post-test are analyzed through Consequentialist Theory, around their awareness of these ethical concerns. Predicted results outline participants’ lack of awareness, and a call to action, attending to the simultaneous act of dismantling racial injustice and constructing educational possibilities.

Author

• Devery J. Rodgers, California State University - Long Beach