Companies improve their business performance by collecting, analyzing, and taking data driven decisions. This is about the “what”, but what happens with the ethical “how” that derives from the usage of data?
In the last decade, the boom of data management and data analytics in the market, has brought almost everyone closer to data ethics in their day-to-day work. From our personal reports to company’s reviews, we use frequently data analytics. Here arises the question: Is the way we analyze data, biased towards some specific social groups? Do we behave open-mindedly when setting the norms of data analysis or we are so attached to our own pace of thought that we cannot react more objectively?
Data analysis is everywhere and thus its ethical part. Humans and their interactions follow the rule of complexity and utilitarianism. Imagine adding to this equation ethics and the power of data. Boom! Gordian knot. Everything becomes more intense on the grounds that even lawmakers are still observing to create regulations in the human – tech interaction. It is a “fresh” necessity which is not yet fully regulated. So, if it is so difficult to regulate norms, imagine what mistakes can be made when working in fast and standardized mode.
Data analysis ethics are part of the overall data ethics guidelines and should follow rules of privacy, security, and transparency. Even if these three rules are easily understood, there is a gap in real case execution as intentions play their part. In most cases, we are following our feeling of rightness. Unless you are a professional in data ethics, you do not always know exactly how to behave towards the right or wrong and you base your activity on your human understanding of rightness. But have you ever thought that the social status/color/gender of the professional that creates an algorithm for data analytics is highly correlated with the outcome of the analysis? Right answer depends on the right question but what happens when the “rightness” depends on the subjective eyes of the beholder? Does this mean that the excellent professional as far as the productivity is concerned, may not be that excellent if his/her understanding is not that holistic? Does this create a threat to the overall business and its customer success if our employees have a limited or malevolent understanding of the world?
“However, even when intentions are good, the outcome of data analysis can cause inadvertent harm to individuals or groups of people. This is called a disparate impact, which is outlined in the Civil Rights Act as unlawful.
In Data Science Principles, Harvard Professor Latanya Sweeney provides an example of disparate impact. When Sweeney searched for her name online, an advertisement came up that read, “Latanya Sweeney, Arrested?” She had not been arrested, so this was strange. “What names, if you search them, come up with arrest ads?” Sweeney asks in the course. “What I found was that if your name was given more often to a black baby than to a white baby, your name was 80 percent more likely get an ad saying you had been arrested.””1
When talking about the “how”, we are talking about the quality and thus it is always difficult to normalize. Data is a treasure we hold in our hands, but it is highly important that data management and analytics departments invest more in data ethics direction of their teams. Our moral responsibility as human beings should rule our professional identity and not be separated. By learning lifelong, supporting the power of knowledge internally and aligning with regulations, companies can make major steps towards a more conscious data ethics direction.
Andromachi Orfanidou, Business Development Representative