These days just about anybody with a computer can gin up an infographic to showcase whatever dataset they so chose to lead unsuspecting readers to predetermined conclusions the creator so chooses. That is a corrupting influence that can lead to very poor decisions and very poor outcomes. To prevent this, a set of professional ethics is needed for data science professionals, akin to the Engineering Code of Ethics or a Doctor’s Hippocratic Oath. I submit that data professionals must adhere to a strict oath that commits them to accurately and precisely collect, analyze, and present data regardless of outcome or the desires of their employer. Specifically with respect to three phases of data science including:
- Data Collection: To use truly non-biased data collection methodologies to accurately represent populations. To utilize truly random populations so as not to skew results in favor of a certain desired outcome.
- Data Analysis: To apply honest and open computational and analysis methodologies in non-biased manner so as to not corrupt the fundamental underlying data.
- Data Presentation: To present data in such a way that is clear, accurate of results, and informs users of the process to collect and analyze the data. To as much as possible not present conclusions but to rather allow the reader of the data draw their own conclusions.
Graphics provided by openclipart.org