Sponsored post by Melbourne Business School (MBS), a partner of IAPA. Prepared by Associate Professor Jennifer George, Director of the Master of Business Analytics at MBS.
In April I had the opportunity to attend a meeting of directors of Business Analytics programs from around the world. The event was sponsored by SAS and held at their global headquarters in Cary, North Carolina. Along with insightful presentations by industry leaders in analytics, I found this meeting to be a good chance to survey the state of analytics education.
The “data scientist unicorn” is a label often used in North America for the unusual set of skills needed by well-rounded analytics professionals in organisations. With a requirement for excellent programming and data warehousing skills, deep statistical modelling and data mining ability, optimisation and mathematical modelling, the ability to understand the business or organisational context of a problem and communicate effectively in verbal, visual and written forms, it is no wonder that these people are as elusive as the unicorn they are named for.
Universities are beginning to design programs aimed at this market. In the past few years there has been a remarkable expansion of analytics degrees around the world, primarily at the master’s level, and these have taken a number of different forms:
- The Master’s program in an IT or Computer Science department.
Often titled “Data Science”, typically these programs major on the data management aspects of analytics. Analytical techniques tend towards data mining and unsupervised machine learning. Statistical and mathematical modelling, communication and business context may be relatively weaker in these programs although Data Visualisation is often well covered.
- The Master’s program in a Statistics department.
Typically these programs major on the statistical or econometric analysis aspects of analytics. Some are simply Master of Statistics degrees that have been rebadged. They are often rather weak on the programming and data warehousing skills required and are usually lacking in the optimisation and mathematical modelling, business context and communication aspects of the analytics professional’s skill set.
- The Master’s program in a Business school or Commerce faculty.
Usually titled “Business Analytics”, these programs often lack the programming and data warehousing skills required but usually have strengths in business context and communication while doing a good job of covering mathematical and statistical modelling and to a lesser extent data mining and machine learning.
- A few programs that try to do it all
The Master of Science in Analytics from North Carolina State University is in its ninth year and so is one of the oldest analytics degrees in the world. The NC State degree is an intensive one year experience, run by an institute in the university established specifically for this purpose, which is fulltime and requires students to commit to a 9 to 5 working day and regular no-holds-barred peer feedback. MBS referenced this model for the program we designed because it addresses all the skills with a well-designed curriculum, practical projects with real company data, teamwork and leadership experiences and a focus on both the computing aspects of data science and the mathematical techniques of optimisation, statistical modelling and data mining. This model requires genuine collaboration across disciplinary silos and time-intensive individualised coaching and mentoring – a pursuit MBS is proud to deliver on. You can find out more about the MBS program online.
As the need for data scientists continues to grow, so will educational programs. Programs that focus on some of the skills will produce graduates who are useful working in analytics teams and this will certainly fill some market demand. However if universities are to truly step up to the challenge of this field then more interdisciplinary programs that bridge all the skills are needed.