This summer I completed my Master's of Science in Data Analytics through Franklin University, and I'm already reaping benefits to my career with a recent promotion to become a Wizard on the Innovation and Excellence team at Leading EDJE. But is a data science related master's degree right for everyone?

Let's talk about my experience in the program, my reasons for pursuing it, and who a data science master's degree would be right for.

Why I pursued a Master's Degree

A number of years ago my father's health declined suddenly and we lost him to cancer detected just 6 weeks earlier. My dad and I had been close, and he'd helped me discover programming and even artificial intelligence as a kid. I was devastated by the loss, and over the course of the following months turned to fun game development courses on Coursera as a way of processing that grief since very little brought joy in the immediate months following that loss.

I eventually ran out of game development courses to take, so I switched to courses on Python and AI, recalling how much I enjoyed AI in my teens and twenties. These AI / ML courses brought a lot of joy and passion to my life and I enjoyed learning these topics more formally. For months I continued to burn through 1 - 3 Coursera courses each week, until my wife finally told me:

These courses sound harder than what I did for my master's! Go get a master's degree already!

So I did.

There's more to it than that: I was teaching at the bootcamp level at the time and wasn't fully sure I trusted the organization after a change in leadership occurred. I enjoyed teaching and wanted alternatives to continue to teach should I leave working for that employer, and a master's degree would do that. I also had benefits that helped pay for the master's degree, making it less of a financial investment on my part. Additionally, there were a number of options for online evening learning that I could do while still working full time, so it was a nice self-paced program that fit my needs. I also was intensely interested in learning machine learning more formally and addressing gaps I identified.

In the end, I pursued a Master's in data analytics for three main reasons:

• As an insurance policy in case I ever moved on from my employer and wanted to teach elsewhere
• For targeted skills development in the area of machine learning
• For personal fulfillment and enjoyment as I knew I enjoyed learning about these things and applying myself

Selecting a Program

I looked at a number of online options for a master's program before settling on 3 program I applied to. I applied to University of Illinois: Urbana / Champagne (UIUC), Franklin University here in Columbus, Ohio, and a college I'll not name in a devilishly hot southwestern state.

I was accepted to UIUC and Franklin, but got a rejection from my third option. Ironically, I was rejected because although I took and passed a math course during undergrad, because when I transferred colleges and changed from a Computer Science major to a Computer Information Systems major that math course wasn't needed, it technically didn't transfer. The school noted that I had taken and passed the course, and that it was on my transcript, but because it wasn't associated with my final undergrad degree, they didn't want to accept me.

UIUC is a fantastic school, but I was afraid that its intensity level might be too high and it would eat up the time I use for conference speaking, user group organizing, book writing, and other things and that it would constitute a significantly higher stress level, though it would provide a more rigorous education. Additionally, I felt that Franklin was part of my local technical community and potentially an institution I could teach at someday as an adjunct instructor if I wanted to.

For that reason, I enrolled at Franklin University and went through a sequence of 8 courses structured around data analytics, statistics, data visualization, data mining, and machine learning.

My Educational Experience

First of all, I have to say how much I enjoyed my time at Franklin University, and how much I appreciate the program chair of the Master's of Science in Data Analytics program in particular. Dr. Alpay was fantastic to interact with from my early explorations of the master's program until graduation, and I am proud to be associated with a program she chairs.

Secondly, I want to stress that educations are not a one-size-fits-all product. For example, my education contained a course on database management systems and a course that was an effectively an introduction to programming with Python in a Jupyter Notebooks environment. I actually already was professionally teaching a significant amount of the content of my database course at the bootcamp level, and the day before my intro to Python class started I gave a user group presentation that wound up containing the same amount of content that the entire upcoming semester was about to cover.

These could have been easy "coast through it" classes, but instead I chose to push myself a bit more. For every week in the 12 week course on database management systems I read a different book on related database topics. While we were studying different notation schemas for relational databases, I was reading about database indexing strategies and how databases work internally. I also was able to take that course's final project and use it to research innovations in Azure Synapse Analytics and deepen my understanding there.

My intro to Python class was even more extreme. I reached out to the professor and explained my situation. With his approval I then submitted all of the homework for the class in the first week, then skipped all remaining lectures and instead independently worked on a computer vision system in Python. Over the course of that class I trained two machine learning models: one to recognize hand gestures in images as "rock", "paper", or "scissors", and another to predict what gesture I was likely to use next based on my past patterns of selecting moves. By the time that course was done, I had an opponent who could fairly consistently defeat me at rock / paper / scissors. This was particularly helpful as I was just starting development of my LinkedIn Learning course on Computer Vision.

Other classes were more challenging - particularly those dealing with heavy mathematical issues. It'd been 20 years since I was last in the classroom and mathematical notation and topics like matrix multiplication were difficult and took me a significant amount of practice to read and internalize.

My favorite courses were those that gave me large projects that I could use to push myself to my fullest extent. My least favorites offered little flexibility, had heavy plagiarism and cheating from peers in the course, or forced us into group projects (I've got 20 years of industry experience and even experience managing and teaching, I don't need group work to teach me how to work in a group!).

The culmination of the master's in data analytics experience was the capstone project which I completed this summer. I was able to pick my own problem to solve and technologies to use, with some added guard rails to make sure I was working within the confines of my professor's expectations. I chose to train a machine learning model to classify code commits as bugfix or non-bugfix related based on the commit metadata and commit messages. I was able to use ML.NET, generative AI, Python, Azure Machine Learning, and Polyglot Notebooks to achieve my results and I'm quite proud of that project. If you're curious, you can read my bugfix classification project write up on my ML.NET blog, but a sample visual generated from the output of that project is pictured here:

Overall, I finished in a 2.5 years and maintained a 4.0 GPA throughout my experience. I also was able to pursue and accomplish significant achievements in addition to my academic studies including authoring the technical books Refactoring with C# and Data Science in .NET with Polyglot Notebooks, creating my LinkedIn Learning course (and a second I'm currently working on), being awarded two Microsoft MVP awards, and continuing to be able to speak at conferences and organize the Central Ohio .NET Developers Group.

Career Impacts of a Data Science Master's Degree

So, was a master's in a data science related degree worth it?

From an emotional fulfillment perspective: yes! I really enjoyed my studies and my ability to push myself and get deep. In all but my first machine learning class I was able to take assignments and right-size them for my own personal learning goals and push myself to learn many new things. That alone probably wasn't worth the roughly $3,000 / course cost, however.

On a skills development front, I am now capable of so much more than I was just a few years ago through applied studies inside of and outside of my collegiate studies. My confidence in my ability to perform, learn, grow, and teach is also significantly higher and I'm finding myself not doing the little things I did before to discount my knowledge or experience level.

But the big question is if it's worth it from a career perspective.

A data science (in my case data analytics) master's degree isn't required for many data science roles, but it can be helpful for persuading someone to give you a chance to get into a data science role for the first time. Experienced data professionals with a related master's also find it easier to have conversations about higher pay bands as well, which can make a difference.

In spring of 2023 I left teaching to return to the software engineering industry as a consultant. My goal was to find an organization I could be a part of as a full-time employee and help them help their clients with software engineering, AI, and machine learning needs. With two decades of experience in software engineering I was very qualified for this role, but the conversation got even easier because I had just earned a graduate certificate in data analytics at the halfway point of my data analytics master's degree (Franklin has certificates like this built into a number of their degree programs).

Having formal credentials, in this case a graduate certificate and a Microsoft MVP in AI, made it easy for Leading EDJE to trust me based on my software engineering reputation and my AI / ML credentials - even though I had not worked in a data science or AI role before that point.

Three months later I had my 90 day review and was promoted to the organization's first AI Specialist. A year later, I've now been able to apply these AI skills in helping my clients pursue their software engineering, AI, and ML goals and help guide them to better solutions. I've also been able to train others internally and continue to make an external impact through teaching at conferences and user groups and through book and course projects.

Just this week we announced my promotion to the role of Wizard, our name for a member of Leading EDJE's Innovation and Excellence team. In my new role I'll be able to help the organization at the strategic level through internal training and external relationship building as I help identify and communicate new trends and practices in our industries and markets and help generate additional leads for the organization.

By the way, if you feel your organization would benefit from having some software engineering, AI, ML, or DevOps consulting, please get in touch!

Was a Master's worth it for an AI and Machine Learning professional?

All told, my master's in data analytics cost around $25,000 USD and took a portion of my time and focus for two and a half years of my life.

While I don't want to discuss salary details or raises, I feel the amount of compensation adjustment afforded by this degree will be at or above that amount long-term.

More importantly, my ability to execute on more complex problems is significantly higher, as is my comfort level and credibility in doing so. Because I supplemented my education with additional learning opportunities, I feel I can continue to grow through targeted studies related to topics of interest and new industry trends that emerge in the coming years.

Additionally, that master's degree makes it easier for my employer to communicate my experience with prospective clients and opens the door to being an adjunct instructor in the evenings later in my life if I want to - as well as future projects such as books, courses, or collaborations.

A master's degree isn't a necessity for all data professionals and won't be right for every individual, but I think it was the right move for me.

Should you get a master's degree in a data science related program?

I've talked about my own journey so let's take a step back and talk about you and if a master's program is a good option for you.

I'd recommend a master's in data science, master's in data analytics, or master's in computer science with a focus in AI if you meet the a few of following criteria:

• You have time for the program - both in terms of the years it will take and the schedule impact of your studies
• You have an alternative source of income during your education and do not plan on taking on new student loans
• You believe you can learn effectively in an academic setting
• You are willing to go above and beyond the materials to make your education practical and relevant to your individual interests
• You expect to receive a significant raise or unlock key new opportunities from achieving the degree
• You plan on teaching at the collegiate level
• You plan on moving on to get a doctoral degree

Not all of these need to be true or completely true for you for an advanced education to make sense for you, and if you're just looking for skills development there are cheaper alternatives such as independent learning, books, and online courses. However, formal study in an educational system has its merit, particularly in forcing a well-rounded education and providing a recognized credential to current and future employers.

I'd love to hear about your own educational considerations and what's important to you from a skills development perspective.