By Russell Willerton | STC Member
We know that information is increasingly valuable. Technology companies collect it from us when we use or buy their products. Marketing companies collect, buy, and sell information about what we buy and when, where, and how we buy it. Your organization, whatever it is, collects, stores, and uses information from those with whom it interacts.
Data collection and data storage are woven into the goods and services offered by companies and nonprofits large and small. When all goes well, the experience is seamless, but when data systems are breached and information is stolen, we are reminded that the information we value may turn dangerous when it falls into the wrong hands. Hackers use stolen information to steal identities and the financial accounts connected to them. Criminals use software to take over personal computers or organizational data servers and hold the contents hostage until ransom is paid.
Because the potential for data to be misused is so great, it is increasingly important to handle data ethically. A recent e-book, Ethics and Data Science, encourages a broad view of ethics when it comes to collecting, sharing, interpreting, and visualizing data. This resource is not only for data scientists but is broadly useful for anyone working with information or data from other people. The authors are professionals with a range of experience in corporate settings. Mike Loukides is a Vice President at O’Reilly Media, Hilary Mason is a data scientist with experience in machine learning and artificial intelligence, and DJ Patil served as the US government’s first chief data scientist.
Why is data ethics something technical communicators should care about? Modern organizations are built around data. The work that technical communicators do—as writers, as user experience experts, as business analysts, as developers of training materials—supports each organization’s uses of data. Technical communicators can advocate for products, practices, and services that manage data responsibly. In this way, technical communicators protect the interests of their users—and their employers.
Below I summarize some key points from Loukides, Mason, and Patil’s e-book.
Use Checklists to Build Ethics into Practice
Many professions have codes of ethics that outline principles of positive behavior. The Hippocratic Oath, or some variation of it, is frequently affirmed by those licensed to practice as physicians. In practice, however, swearing an oath tends to occur once or infrequently, and provisions for enforcing oaths are few. Similarly, many professional codes of conduct or codes of ethics are difficult to enforce, although some professionals, like doctors or lawyers, may lose their privileges after particularly egregious lapses in ethics.
However, checklists can affect activities positively if they are built into an organization’s procedures and practices. Loukides, Mason, and Patil point to surgeon Atul Gawande’s book, The Checklist Manifesto: How to Get Things Right (a favorite of mine), as a positive example. Gawande provides data from around the world showing that hospitals big and small had better outcomes and fewer deaths after surgeries when members of the surgery team followed checklists for safety and hygiene. An important aspect of the surgery checklists is related to enforcement: Without unanimous consent that a checklist item has been satisfied, the surgery cannot go forward. This means the most politically powerful person in the room—the surgeon—cannot simply boss around the others and begin an operation before the team is ready.
Similarly, the authors point to the principle of the “Andon Cord” in the Toyota Kaizen method of manufacturing. Any worker on a production line may pull the cord, figuratively or literally, to stop the line if a problem arises. Contemporary companies frequently develop new products quickly, pushing them out and then reacting when they break, fail, or go awry. If a company culture values data ethics, employees will be empowered to raise concerns and take action when data use problems arise instead of merely hoping that they go away.
Technical communicators frequently use checklists for audience analysis, project planning, and project management. Loukides, Mason, and Patil offer several checklist items that can be used to generate discussion about ways to collect and use data ethically. Discussion of ethics can be integrated into just about any process. Of course, checking a box does not ensure that a person or group is behaving ethically. As companies integrate ethical concerns into their processes, they also integrate ethics into the corporate culture.
The Five Cs of Data Ethics
Loukides, Mason, and Patil provide the following five concepts, or framing guidelines, to consider when collecting and using data from users.
- Consent: Companies should not collect data from users of a product or service without users’ consent.
- Clarity: Terms of service should clearly state what a company wants to do with the data collected; too often, these documents bury a company’s intentions behind obtuse, confusing verbiage.
- Control and transparency: Related to clarity, companies should give users levels of control over their data instead of forcing all-or-nothing choices.
- Consistency and trust: Companies should demonstrate their concern for users by consistently and proactively protecting their data.
- Consequences: Companies should protect against unforeseen consequences of data loss or misuse—and even decide against products and services that involve great risk. One example of unforeseen consequences can be found in the case of Strava, which makes wearable GPS devices. When the company released data to help people find new locations for exercise, the locations of US troops wearing Strava devices in Iraq and Afghanistan were revealed.
Connecting Ethics to Company Culture
Loukides, Mason, and Patil emphasize that abstract ideas about ethics need to become part of a company’s culture. They list several opportunities for this, including asking prospective employees to respond to questions involving ethical issues and going beyond standard financial concerns to question a new product’s impacts in environmental, cultural, and ethical areas. To promote further discussion, they provide a link to case studies on ethics and artificial intelligence from Princeton’s University Center for Human Values (UCHV) and Center for Information Technology Policy (CITP).
Take an inventory of your company’s ethics: What principles of data ethics does your company uphold? To what degree is ethics a part of corporate culture? What steps does your organization take to protect customers’ data?
Loukides, Mason, and Patil point out that strong ethics are good for business. If you see an opportunity to improve your company’s data ethics, don’t let it pass by.
Gawande A. The Checklist Manifesto: How to Get Things Right. New York: Metropolitan Books; 2009.
Loukides M, Mason H, Patil DJ. Ethics and Data Science. Sebastopol, CA: O’Reilly Media; 2018.
Princeton Dialogues on AI and Ethics case studies. Princeton Dialogues on AI and Ethics website. https://aiethics.princeton.edu/case-studies/