Master of Data Science and Analytics has become quite well implemented as part of universities’ curriculums. They are now over a thousand masters offered across different departments and schools, from schools of Computer Science, Mathematics & Statistics, Economics or Business.

When considering a MSc in Data Science, a few questions will probably pop up in your mind:

• Do I need a Master's in Data Science or Analytics to become a Data Scientist?
• Do I need to follow a Master of Data Science or Analytics to learn Data Science?
• Will a Master of Data Science or Analytics Help me get a job in the field?
• Are they differences in the available masters out there?
• Does a Master of Data Science provide a good return on investment?

Do I need a Master's in Data Science or Analytics to become a Data Scientist?

The short answer to this is no. A lot of Data Scientist comes from a wide range of background. In her 2016 Forbes article, Meta Brown described that the number 1 reason for not getting a Master of Data Science is “that you can get a Data Science job without getting an additional Degree.”

At the initial stage of Data Science, having a Ph.D. or at the very least a Master's degree was pretty much required (some sources indicate close to 80% of data scientists with a graduate degree in 2015). these days this is shifting, and most Data Scientists have only a Bachelor’s degree.

Data Science is overall in need of quite advanced analytical, numerical, and computer science skills. However, there has been quite some advance in Machine Learning tooling, helping democratize a bit the area. This advance has been coupled with education programs that have been more in line with the need for Data Science, namely increased computer science education and increased importance of stats/ml components as part of bachelor programs.

These days the same way as it is not uncommon to find a Master of Data Science, there are programs at the bachelor's level that are already focused on the field.

Nevertheless, an academic Education in Data Science is not required to be a Data Scientist. Many other programs provide academic foundations that make it reasonably easy to transition into Data Science with some self-learning.

Some engineering programs, for instance, already cover quite a lot of the materials in terms of quantitative, stats, and programming elements and might even integrate some typical Data Science courses as options. The field of Data Science tends to absorb topics that have been traditionally covered in other subjects.

This is the case for causal methods; these methods traditionally come from Econometrics and have now been adopted in Data Science. In the same way, Data Science has been embracing Econometrics methods; Econometrics has been adopting Data Science methods in an effort to modernize, making the frontier between the two fields even blurrier. Joshua Angrist, the author of one of the most famous Graduate Econometrics textbooks, “Mostly Harmless Econometrics,” and the 2021 Nobel Prize in Economics, describes Econometrics as the original data science.

Hal Varian, the chief economist at Google, was one of the people who has helped popularized Data Science through his famous quote in the 2012 Harvard Business review article “Data Science is the Sexiest Job of the 21st Century”, with his famous quote “ “The sexy job in the next 10 years will be statisticians.”. Hal Varian’s work at Google at the time was heavily focused on ad Auction, something that these days most would associate with Data Science — but utterly grounded in Economics theory.

Economics and Econometrics are not the only fields with deep ties to Data Science, at another hand of the spectrum, the neuroscience field had a profound impact on shaping some aspects of Data Science, particularly in its more quantitative and computational incarnation. Computer vision for instance has benefited tremendously from the cross-pollination between computer science and neuroscience and their contribution to the field of deep learning.

Behind the Data Science term, what is present behind the scenes is a constellation of different methods centered around analytical, statistical, and computer science capabilities to process Data. There isn’t just one way to become a Data Scientist. Still, ultimately, data scientists can come with a different domain knowledge/ academic background, focusing on traditional prediction methods, time series, causal methods, bayesian, deep learning, and experimentation …

A Data Science degree provides one pathway to a job as a Data Scientist, but it is far from the only one.

Do I need to follow a Master of Data Science or Analytics to learn Data Science?

Larrabee’s Law is defined as “ Half of everything you hear in a classroom is crap. Education is figuring out which half is which.”. Could a more focused education save be more effective then?

There are many different paths to learning Data Science. From self-learning to going through specific certificates, to going through specialized Data Science boot camps.

Data Science Masters typically have a few academic goals — these vary depending on the specific Master at hand. Some will try to prepare you academically for further studies/research. Some of which might try to prepare you for particular roles in the job market. Others might prepare you to increase the chance of landing a good job. Some might attempt to do all this at once.

The vast majority of Data Science roles do not require such a wide array of skills, deep understanding, or deep background knowledge, perhaps in the fields of mathematics or statistics. There is, therefore, quite a large amount of knowledge learned that is practically overhead when undergoing an MSc in Data Science.

Looking at my Alma Mater’s Bocconi University, Data Science MSc and its’ study plan’s content. There are very few courses (14 out of 120 credits) that had direct application to the different roles I had as a Data Scientist; on top of that, about 16 credits I would have considered as the foundation for these courses, so at most the equivalent of 1 semester worth of learning. The other courses provide breadth and enhanced academic foundations.

There is, however, quite some overlap in terms of foundation knowledge with other courses. When I underwent my studies, an MSc in Data Science was not ordinary at the time. However, my university did not offer one — they did offer a quantitative MSc in Economics—covering part of the same foundational courses as in their Data Science MSc — yielding less than 1 entire semester of self-study needed to acquire the knowledge needed to operate as a Data Scientist.

Many of the content provided in these academic courses are often somewhat disconnected from the industry's needs. As much as the theoretical knowledge taught in some of the courses, practical experience in cleaning dirty data, data extraction, and modeling based on real-data experience are necessary. How to deal with these is usually not taught in university and needs to be learned from practical experience — on the job.

Overall, going through a Data Science MSc helps learn Data Science, but it isn’t the only way. Other MSc and Bachelor's degrees provide sound foundations for learning the craft, and a lot can be self-studied if there are the right quantitative foundations.

Will a Master of Data Science or Analytics help me get a job in the field?

About half of Data Scientists these days have a master's or Ph.D. degree. Having a graduate degree helps to be competitive with the academic qualifications of your peers when candidating for a job opening.

Furthermore, some countries’ educational systems tend to have it as a standard to have a Master's degree for most educated white-collar jobs. If there is already a need for a master's degree, if you are attempting to get into the field, you might as well choose a relevant master's degree.

In 2022, the demand for Data Scientists is still quite strong and Data Science and Analytics masters are quite new. The requirements from hiring managers do not necessarily tend to include a Master's degree in the field as mandatory. Most hiring managers tend to be quite open with regard to education — looking more for general quantitative education as well as some evidence of having gone through some programming and machine learning courses.

Are they differences in the available masters out there?

There are significant differences in course content between the different masters offered. A lot has to do with the department or school’s history.

Quite a number of the Data Science and Analytics Masters are essentially existing masters slightly altered and re-branded as MSc in Analytics or Data Science. For instance, this is the case for the MSc in Business Analytics at Lancaster University, which is essentially their pre-existing MSc in Operational Research and Management Science re-branded.

Masters of Analytics and Data Sciences, being offered at a wide variety of departments, from computer science to statistics, to economics to even the likes of the Health Department, will push the curriculum in different directions. Some might have a bit more focused on deep learning technologies and applications such as NLP or computer vision, some attempt to bring about more domain focus, and some might be more focused on the programming aspects of statistical theory.

This difference in master's might also show in the entrance requirements of the different programs. Some might require a bachelor's in Computer Science, others in Mathematics or economics.

Does a Master of Data Science provide a good Return on Investment (ROI)?

Besides providing instruction, one of the main goals of undertaking a degree is the concept of Educational Signalling, the idea is that the degree shows that you have certain knowledge and competence. The concept of signaling is particularly used to justify attending expensive elite institutions. Josh Angrist, the 2021 Nobel of economics, studied labor economics, one of his studies focused on the impact of education on income; he states his results as:

Besides providing instruction, one of the main goals of undertaking a degree is the concept of Educational Signalling, the idea is that the degree shows that you have certain knowledge and competence. The concept of signaling is particularly used to justify attending expensive elite institutions. Josh Angrist, the 2021 Nobel of economics, studied labor economics, one of his studies focused on the impact of education on income; he states his results as:

For a field such as Data Science, where it is perfectly possible to learn the base of the craft using free online resources, the question of whether it is worth paying tuition fees and spending 1 to 2 years of your life.

To understand whether it is worth attending these MSc, two things are necessary 1) An overview of the cost of attendance. 2) understanding the different benefits of taking part in these masters.

On the cost of attendance, it can be broken into different subcomponents 1) Tuition fees, 2) Board and Lodging, 3) Opportunity cost (ex-Lost Wages) 4) time

Cost of attendance

The first one is the location and international student status. There are significant differences in tuition fees across master's programs in Data Science and Analytics. Master programs in some locations, such as the US or Australia, have a significantly higher cost than those offered in Europe; for instance — Education might even be Free, such as in Finland or Denmark, where European citizens don’t have to pay any tuition fees.

European countries typically offer some subsidies and might have set prices for degrees (e.g., Public universities in France and Belgium).

Within countries, there can be wide variation in how much these Masters costs.

In the United Kingdom, most courses fall within the £5–£15k range for locals, but outliers exist. These tend to be courses offered by business and economics departments or schools, such as the Warwick business school (£26,000), London School of Economics (£31,584), Imperial College (£34,300), Kings’ College (£35,500), London Business school (£39,200).

I will not delve into the cost of board and lodging or opportunity costs. These are not specific to an MSc in Data Science or Analytics but more to do with location or individual circumstances. They nevertheless need to be considered when deciding whether to undergo one of these MSc. They are also particularly influenced by another factor — time or length of the program.

These degree programs tend to be primarily offered in the 12 and 24 months variety, with a few programs at the 18-month mark and some beyond 24 months. The programs going over 24 months of duration are typically part-time programs.

Benefits from undertaking these masters

There are typically many benefits from undertaking a Master's program.

On the front of career opportunities, it does signal that you have gone through a certain level of education and can probably solve some level of analytical challenges. It also provides students with the ability to benefit from the university's alumni network and career service and their connections to different companies.

In many countries, a master's degree also opens the door to a number of public service jobs, which require a set education level. Furthermore, the weight of education in the selection criteria of some countries is still quite strong, with employers expecting candidates to have at least a two years master's under their belt.

These master's degrees also provide an easier way to transition your career to analytics or data science if you were in a different profession.

Another aspect to consider is immigration opportunities. Many countries offer those who study within the country a temporary “training” visa after their studies. For instance, this is the case in the United States, which typically offers a one-year (OPT) after studies. However, there is a two-year extension for this program available for students having undergone a STEM education. This is a classification offered to most of the Masters of Data Science or Analytics in the US.

Summary

MSc in Data Science or Analytics is not necessary to get a job in data science or to learn analytics or Data Science, there are many other pathways.

They do however provide a way to learn some of the content for Data Science and offer a number of benefits. They provide an academic credential that can open some doors, they as well provide a pathway to immigration in some cases (e.g. OPT US).

If you want to do a Master's degree and get into the field of Data Science, it is not a bad option to go for one of these courses.

‍