Integration of Soft Skills in Data Science Courses

Dr. Ceni Babaoglu, Assistant Program Director, Data Science, Toronto Metropolitan University

Dr. Ceni Babaoglu, Assistant Program Director, Data Science, Toronto Metropolitan University

Soft skills, regardless of graduate or undergraduate student, are key success factors in the data science career and lifelong learning. I advocate the need to teach soft skills as a key ingredient of the rigorous training chores that data science students have to undertake. My utmost effort is to introduce these skills to the students based on my research on data science-related job requirements, hiring practices, and information on the roles and responsibilities of data scientists, engineers, and analysts. Here my emphasis is on communication, critical thinking and problem solving as highly valued skills for all credentials be it a certificate, undergraduate or graduate degree in Data Science, or related field. This article elaborates on the integration of these soft skills in my teaching methodology.

The first step towards embedding soft skills in coursework is to provide a clear and meaningful synopsis to students of how each soft skill is exemplified, practiced, and evaluated throughout the course assignments, projects, and deliverables. I employ role models and exemplars from industry and alumni graduates, personal experience, and job market research to motivate my students to give equal importance to the mastery of soft skills as to the depth and breadth of data management and analysis skills.

The foremost vital communication skill in data science is to effectively observe, read, listen and understand the context, implications, and relevance of the business problem.  The ability to understand business requirements and goals and transform them into implementation details of a data-driven solution is a key factor that remains external and overlooked amid the myriad of learning activities. Here, I emphasize the importance of communication skills to engage in a constructive dialogue with the stakeholders. The academic learning activities need the incorporation of opportunities where students get authentic exposure to represent, express, and engage with clients, regulators, and practitioners. I embed this exposure in my courses through a self-sustaining peer network realized through invited talks of alumni and seminars of industry leaders. Secondly, verbal and written communication skill is required to present the data stories, implications, inferences, and characterization of machine learning models to audiences of different levels. Presenting capstone as well as graduate degree research projects is one key mechanism that I practice to allow students to perfect their verbal communication skills. The end-to-end execution of course projects prepares students to gain and demonstrate foundational project proposal writing skills as well as get exposure to writing research articles along with analytical skills.

" The foremost vital communication skill in data science is to effectively observe, read, listen and understand the context, implications, and relevance of the business problem "

Data science professionals need to evaluate data, information, and domain knowledge in a systematic, efficient, and effective manner that requires critical thinking and problem-solving. They have to constantly innovate and find alternative solutions or methods in the implementation, deployment, and monitoring of the data-driven systems. This makes critical thinking and problem-solving very important skills that students must exercise. Involving students in industry collaboration projects provides ample opportunities to gain this capability. Like any other applied field, the pedagogical treatment of data-driven problems is quite different from a real-world application. Without gaining first-hand experience of getting involved in solving industry problems, students will not learn to ask the correct questions, relate the context with findings, and think out of the box. One effective approach is to base the culminating assignments such as capstone projects and research projects on industry-sourced datasets and problems.

The data science landscape is cross-disciplinary that requires new practitioners and learners to navigate an overwhelming range of concepts, techniques, algorithms, and implementation platforms. Interleaving and connecting theoretical content with real-world business value often become a challenge. Despite gaining an in-depth technical skill set and theoretical aspects of different big-data tools, programming languages, and frameworks, the students feel challenged in mapping their knowledge to a real project. The answer is to make soft skills a part of the teaching methodology and course design.

Data Management

Machine Learning