Data science education in the United States lacks an emphasis on ethics
Undergraduate training for data scientists – dubbed the sexiest job of the 21st century by the Harvard Business Review – fails to prepare students for the ethical use of data science, according to our new study.
Data science is at the crossroads of statistics and computer science applied to a particular field such as astronomy, linguistics, medicine, psychology or sociology. The idea behind this data analysis is to use big data to solve otherwise intractable problems, such as how healthcare providers can create personalized medicine based on a patient’s genes and how companies can make purchase predictions based on customer behavior.
The United States Bureau of Labor Statistics projects 15% growth in data science careers over the period 2019 to 2029, corresponding to increased demand for data science education.
Universities and colleges have responded to the demand by creating new programs or reorganizing existing ones. The number of undergraduate data science programs in the United States has grown from 13 in 2014 to at least 50 in September 2020.
As data science educators and practitioners, we have been prompted by the growth of programs to study what is covered and what is not in undergraduate data science education.
In our study, we compared undergraduate data science programs with the undergraduate data science education expectations offered by the National Academies of Sciences, Engineering and Medicine.
These expectations include training in ethics. We found that most programs devoted considerable coursework to math, statistics, and computer science, but little training in ethical considerations such as confidentiality and systemic bias. Only 50% of the degree programs we studied required ethics courses.
Why is this important
As with any powerful tool, the responsible application of data science requires training in using data science and understanding its impacts. Our results align with previous work which found that little attention was paid to ethics in data science degree programs. This suggests that undergraduate data science programs can produce a workforce without the training and judgment necessary to apply data science methods responsibly.
It is not difficult to find examples of irresponsible use of data science. For example, policing models that have built-in data bias can lead to a high police presence in historically over-policed ââneighborhoods. In another example, the algorithms used by the U.S. healthcare system are biased in a way that results in black patients receiving less care than white patients with similar needs.
We believe that explicit training in ethical practices would better prepare a socially responsible data science workforce.
What is not yet known, and the rest
While data science is a relatively new field – still defined as a discipline – guidelines exist for training undergraduate data science students. These guidelines raise the question: How much training can we expect in an undergraduate degree?
National academies recommend training in 10 areas, including ethical problem solving, communication and data management.
Our work has focused on undergraduate data science degrees at universities ranked R1, which means they engage in high levels of research activity. Further research could examine the amount of training and preparation in various aspects of data science at the master’s and doctoral levels and the nature of undergraduate training in data science at universities of different research levels.
Since many data science programs are new, there is a huge opportunity to compare the training that students receive with the expectations of employers.
We plan to expand our findings by investigating the pressures that may be driving the development of degree programs in other disciplines that are experiencing similar growth in the labor market.
Jeffrey C Oliver is a Data Science Specialist at the University of Arizona, USA. Torbet McNeil is a doctoral candidate in Educational Policy Studies and Practice at the University of Arizona, United States. This article is republished from The conversation under a Creative Commons license. Read it original article.