As our society continues to rely more and more on data, algorithms, and analytics, it’s becoming increasingly important for those within the technical disciplines to be conscious of the many ethical implications at stake. And when it comes to PLNU’s computer and data science students — and those in related disciplines — it’s no different.

This is the thinking of PLNU professor of computer science Lori Carter, Ph.D. While computing students at PLNU have always been required to take an ethics course, they didn’t always find it to be all that relevant to their specific discipline — it often seemed untethered from practical application.

“A few years ago students were telling me that the ethics course they were required to take was very abstract and they didn’t understand how it applied to the things they were learning in computer science classes. It was siloed and abstract,” Carter said. “And so about three years ago I was at a conference and heard someone speak about the much larger scope of ethical issues when it comes to computer science, including issues of diversity, accessibility, and mental health.”

This gave Carter the idea to revamp the curriculum at PLNU and to start developing a module-based education model that integrates ethics in more practical and relevant ways for computer and data science students.

While issues related to computer science like hacking and identity theft pose obvious ethical concerns, Carter is also concerned about the danger of the misuse of data to misrepresent, underserve, or harm certain demographics. In other words, there are major ethical implications with how data is reviewed, analyzed, and used to effect change in the areas of finance, healthcare, business operations, digital communications, product development, and more.  

Carter provided an example to clarify what this might look like. Suppose we’ve gathered data from a survey for college students living on campus in order to determine the reasons why some drop out of school early. Let’s assume some of the students surveyed are females who left because they were pregnant. In this case, some of these women might not be comfortable sharing this personal information in the survey and may leave certain responses blank as a result. 

Those collecting and analyzing this data as a result of the survey, those tasked with “cleaning it,” will now have to determine how to handle these non-responses. Do they simply remove all female participants from the data that didn’t respond to all questions? Do they try to account for the non-responses within the data? Do they analyze multiple data sets and notate how they cleaned the data in each one?

What they do is critical because whatever the data scientists decide to do could greatly change the results and findings, consequently influencing the alleged reasons why college students living on campus dropout and any intended strategies to address them. In the case where these women are simply removed from the data set, for example, this would lead to inaccurate findings that could benefit male college students at the expense of females. 

Carter explains that this same scenario can play out in a number of industries. In the finance industry, for example, banks rely on algorithms dependent on certain data sets that could unjustly keep certain demographics from being eligible for a loan. In healthcare, an area of grave concern during the Coronavirus pandemic, algorithms based on certain data sets are being used to influence the treatment of patients.

“There are predictive algorithms that decide who gets the ventilator and who doesn’t,” Carter said.

Carter also emphasized that it’s important for students interested in pursuing research or academia to understand these ethical issues. Since researchers and academics are often pressured to publish findings in order to build or maintain credibility, this can foment rampant incentives to manipulate data to serve the researcher but lead to misleading, inaccurate, or harmful visual or statistical representations.

“The pressure to publish is so strong, and so sometimes researchers make decisions about data so they can publish,” Carter said, also explaining that data can also be presented through various graphs and visuals in order to mislead.    

“With all of the predictive algorithms we have today are we taking into consideration the bias that could exist? Are we thinking about diverse users when we develop a product? When we create social media platforms that encourage extreme use is that ethical? We felt like as Christians we ought to be the ones leading the efforts to provide ways for our computer science and data science students to have ethics integrated into their curriculum,” Carter said.

Carter, in collaboration with others at PLNU, is in the process of developing a module-based curriculum to meet this pressing need. Instead of having computer and data science students take a single ethics course, they now engage in two or three modules in every computer and data science class they take at PLNU. Each module, which lasts roughly 30 minutes during a single class session, is designed to connect the specific area of study students are learning about to its wider ethical implications.

“The idea is to create these modules that can be easily used and be clearly connected to the course material. The goal is to introduce students to ethical frameworks, from virtue ethics or deontology, for example, so that they have a toolbox to build upon to make ethical decisions,” Carter shared.

The key is to help students think about the work they’re doing through an ethical framework as well as have the language to discuss it with others.

“In a module we may present a messy data set and then ask students what they would do with this data and why,” Carter explained. “The big thing is for them to be able to describe why they did or didn’t do something with the data set.”

When Carter began presenting this module-based curriculum for computer and data science students at conferences both Christian and secular universities showed tremendous interest in adopting something similar for their own students. 

The key is to help students think about the work they’re doing through an ethical framework as well as have the language to discuss it with others.

Although rolling out the modules has required much work, it hasn’t gone unnoticed. In fact, PLNU, in conjunction with Calvin University, received grant money for the new curriculum, further signaling the heightened demand for this type of education for STEM students across the country. 

Carter is grateful for the aid she has received in developing the new curriculum. Fellow PLNU professor Catherine Crockett, Ph.D., has been very helpful. And last summer Carter and Crockett worked closely with two PLNU students, Whitney Featherston (20) and Morgan Wheeler, to run focus groups to determine what students might be receptive to and how PLNU could ensure the modules would be valuable.

As Carter continues to develop and improve upon this module-based curriculum, she is hopeful they will greatly serve PLNU students after they graduate and well into the future.  

“What our goal is for students is to think about the ethical issues of whatever they are involved in, finance or computer equipment or the biological sciences, and to be able to articulate to the people they are working with why they think there is a problem with how we’re doing something,” Carter said. “And this is why we are using the frameworks, so students can speak articulately to their boss or their colleagues about these things.”   

Related Story: The Valiant Work of Cybersecurity