Ethics Code Seeks to Deter Dishonest Analytics
The INFORMS code of ethics requires analytics professionals to report unbiased answers, not just what the client wants to hear.
Wed, February 26, 2014
CIO — When the Institute for Operations Research and the Management Sciences (INFORMS) developed its Certified Analytics Professional certification, the professional society decided that certified individuals should agree to ethical standards as well as meet other requirements. So INFORMS rolled out a code of ethics for analytics last March. As chair of the certification board, Scott Nestler discusses the importance of ethics in analytics.
Why tie ethics to certification?
It can inform the individual about the rules and standards of their profession. It can hold the profession accountable to the public, or at least to their customers and clients. It can help them in their decision-making when they face difficult ethical situations. It can help deter unethical behavior. And it's a way of promoting self-regulation.
What ethical dilemmas do analytics professionals face?
Analytics professionals have obligations to several different groups: to their employers and their clients, to their colleagues, to their professional society, and to society in general. They have a series of obligations there, and it's possible they don't always line up perfectly with each other.
With employers or clients, you have to make sure you're giving them your best professional judgment. And in a client context, you have to clearly state your qualifications to do the work; don't promise more than you can deliver. And analytics professionals are often given privileged information--data to analyze that, if it were to be shared, might cause harm to a business--so it perhaps would be unethical to [take] what you had from company A and use it for company B's advantage.
Sometimes when an employer or a client asks an analyst to look into a problem, it's clear they have an outcome they'd like to see from the analytics project, but the analyst has an obligation to apply the appropriate procedures correctly and report whatever results they find--and not do what might be required to produce the results they think the person who's paying them wants to see.
Do employers pressure analysts to act unethically?
I don't know that it's necessarily overt; it may be more subtle or gray, or it may be that an analyst could look at the situation and think they know what the desired outcome is and think, "If I produce that result I might get hired again."
How could unethical behavior harm the organization sponsoring the work?
Going back to the purpose of most analytics, it's typically to help make better decisions. If you don't get an honest, unbiased, best professional opinion--and realize that most of these things don't have a right answer, but it's a best professional opinion based on the evidence or data--it can cause a decision that can be harmful to the organization, maybe not in the short run, but in the longer run.
It could have financial consequences, or it could have a larger environmental or societal impact. It depends on the organization and the decision they're trying to make. The negative impact depends on how big a decision it is in the first place.