Flickr Photogallery

Subscribe Newsletter

subscribe with FeedBurner

Data Science Education Needed Focus On Ethics

  • September 12, 2021 at 6:56 am
Data Science Education Needed Focus On Ethics

The big idea Our new study shows that undergraduate training for data scientists. Which has been call the most sexy job in the 21st Century by Harvard Business Review, fails to prepare students for ethical science.

Data science is the intersection of statistics and computer science, applied to a specific field like astronomy or linguistics, medicine or psychology, or sociology. This data crunching aims to solve otherwise unsolvable problems. For example, how can health care providers create personalized medicine based upon a patient’s genetics or how can businesses make purchase predictions based off customers’ behaviour.

According to the U.S. Bureau of Labour Statistics, there will be a 15% increase in science jobs over the 2019-2029 period. This corresponds with a higher demand for science training. Colleges and universities have responded to this demand by creating new programs, or rewriting existing programs. In the United States, there are now at least 50 undergraduate data science programs. This is a significant increase from just 13 in 2014.

We were both educators and practitioners of science and we were inspire. By the increase in programs that investigate what and isn’t cover in undergraduate science education.

Our study compared undergraduate data science curriculums with the National Academies of Sciences, Engineering, and Medicine’s expectations for data science training. These expectations also include ethics training. Most programs had a lot of coursework in mathematics, statistics, and computer science but very little instruction in ethical issues like privacy and systemic bias. Only half of the degree programs that we examined required coursework in ethics.

Why Data Is Important

Data science is a powerful tool. Responsible application requires knowledge and training. These results are consistent with previous research that showed little regard for ethics in science degree programs. This suggests that students in undergraduate data science programs might not have the skills and judgement to use science responsibly.

You can easily find examples of reckless data science use. A built-in bias in can cause a high level of police. Presence in neighbourhoods that are historically under-police. Another example is the U.S. healthcare system’s bias algorithms. That allow Black patients to receive lower care than those with similar needs.

Extensive training in ethical practices is a better way to prepare scientists who are socially responsible.

It Isn’t Yet Known What The Future Holds

Data science is still a new field, but guidelines exist to train undergraduate students in the discipline. These guidelines raise the question: What training can an undergraduate get?

The National Academies recommend that you receive training in 10 areas including communication, data management, and ethical problem solving.

Our research focused on undergraduate science programs at schools that are R1 (research-intensive). Further research could be done to examine the preparation and amount of training in data science at Masters and Ph.D. levels, as well as the nature of undergraduate scientist training at schools with different research levels.

There is a lot of opportunity to compare students’ training with employers’ expectations, as many data science programs are still new.

Next Data steps

We will expand our research by exploring the factors that could be driving curriculum development in degrees in other disciplines, which are experiencing similar job market growth.