Data Science Sounds Cool

Three years of undergraduate courses under my belt and I could not choose one. I could go to Ross. The school is well established and my degree would surely hold weight. I could also engage in Philosophy. The debates about diachronic personal identity could keep me pondering all day — until a migraine starts sneaking in.

It wasn’t until I spent several sleepless nights running my code in the terminal and creating thorough test cases to ensure a smooth program that I realized I may have figured out my next step in life. The thought of being a Data Scientist sparks excitement to know that there is a plethora of industries to be a part of and surely the promising salary to support a pleasant lifestyle.

As my knowledge expands more in the field, controversial topics begin to surface — including the many uses of data analytics and how they are obtained. Systems get created to analyze all of society and then conclusions are formed as if “one size fits all”. With this thinking, who gets to determine that the numbers look correct?

In a book discussing how data systems can put some people at a disadvantage while benefiting others, O’Neil found that it was not the system’s fault for producing the numbers it was created to produce — it was the people designing these systems. Every individual who contributed in creating a model that would conclude something about a person inserted their own misunderstanding, prejudice, and bias into the software system which dictated people’s lives [1]. There are, nonetheless, many individuals behind these data systems with good intentions. However, it does not outnumber the many who may not consider aspects of people’s lives holistically.

O’Neil’s book also highlighted a school teacher, Ms. Sarah Wysocki, who was a victim of these inaccurate data systems. She taught in an underprivileged community and helped her students succeed academically. Her school introduced a test that teachers would take annually in order to continue their employment at the school. One year, Wysocki, did not pass the test and because this data system was put in place that deemed teachers incapable of helping students succeed if you do not pass this standardized test, she was out of a job and the school was at a loss of a great teacher. This data system created by individuals affected the life of Wysocki, her students, as well as the school system she was once a part of because an algorithm created to fit people of different backgrounds and experiences into one uniform category.

Wysocki would have still been employed if the idea of the standardized test was not put in place. People contributed to the formation of the system and decided what was a good cut off for teachers and what was not. These certain behaviors of contributing one’s own bias to these algorithms then make me wonder if the values of systemic inequality are rooted solely in individuals or if they are embedded in these data industries. It is concerning because my values will not allow me to practice behaviors that will deeply affect individuals and know that it may be causing damage to some communities. In a world of constant change, I hope to find an industry that aligns with my values and advocates against creating biased data systems.

References

[1] Cathy O’Neil, Weapons of Math Destruction — Crown Publishers, (2016)

[2] Image, https://towardsdatascience.com/survey-d4f168791e57