Diversity in Data Science Team
SEPTEMBER 20, 2017
Just like any other positions in the job market, the qualification for data scientists for most of the companies is congruent and specific to match a certain group of professionals. To be recognized as a candidate, it is almost definite to have a degree in Statistics, Computer Science, or related fields, and years of professional experience is mandatory. Companies use a narrow specific job description in recruitment while expecting their data scientists to know it all from Mathematics, Statistics, machine learning, programming, to business and domain knowledge; it is no different from searching for a unicorn. Why do you ask a machine learning expert to know business better than a business person and expect a statistician to master marketing? And why don’t you have all of them as a team to share their expertise to fill each other’s gaps and add values to your business?
Not just in a data science team, people typically overlook the strength of diversity in workplace. People from different majors may have unique potentials and contribution valuable to the team. Especially in data science which covers a variety of content, business questions, and applications, each process requires varied skills and knowledge. The same data may be interpreted differently, depending on questions and goals. Given a generic example of customer segmentation task for a restaurant chain, we may expect that demographics and socioeconomic status are potential factors but specialists in the field may also take regular diet, social circles, and commuting routes into an account. Similarly, doing the same task for a hospital may use a totally distinct group of factors, such as medical history, family profile, and health concerns.
Gender diversity also matters here as males and females, in general, focuses on different topics and details of the matter and, in turn, provide extensive perspectives on analytics. Research on cognitive and psychological variations in males and females has proven non-negligible differences of their visions, interests, and concerns. Leaving the traditional concept of “Math and Science are for boys” aside, we can notice the benefits of gender equality to broaden perspectives and idea-sharing. Given a case study of campaign effectiveness analysis for a fashion and clothing retailer, it is twice as hard for a male- or female-only team to investigate gender-specific products or campaigns, or try to interpret insights underlying buying behaviors of the opposite gender.
With the variety and complexity of data we are facing in today’s business, it is necessary to have a strong data science team to confront any data challenges. In order to build the all-around team, it is unnecessary to hunt for the so-called data science “unicorn”; it is about how to create an integration of skill and knowledge sharing among team members, who are from diverse academic and professional background, in order to see through every dimension of data and complete the whole picture, like building blocks”