Georgia Tech’s MS Analytics Program: My Review

In late 2017, I started looking at different online masters programs with a heavy focus on quantitative skills. This included statistics, data science, analytics, economics, and various MBA programs. Unfortunately, many of the top programs, such as Carnegie Mellon’s Business Analytics or Berkeley’s Data Science program, had crazy price tags of \$50,000 or more. While a higher salary is not the only reason I wanted a masters, the high tuition put a sizable dent in the potential return on investment. After researching programs for several weeks, I finally found what I was looking for. This analytics program had a price tag under $15,000, had flexibility in different specializations, and came from a reputable university. It was Georgia Tech’s recently formed Online Masters of Science in Analytics program (OMSA).

Fast forward to today and I am currently enrolled in their Financial Modeling course. Upon completion, it will be my 21st credit out of the 36 needed to finish the program. If everything goes right, I am on pace to finish the program by the end of next year. I have taken the conservative approach of enrolling in one course per semester because of my full time job.  If you are interested in a quantitative-heavy online masters program, keep on reading to see my honest review of this program.

Admissions

One of the conveniences for choosing Georgia Tech’s program was the minimal requirements for admission. They did not require a GRE or GMAT score which saved a lot of time in the application process. The lack of requirements can be seen both as a strength AND a weakness. One of my biggest complaints with this program is the huge disconnect between the admission requirements and the expectations in class. The requirements state that you only need at least one college-level course or equivalent knowledge in “Computer programming in Python at the level of Introduction to Computing in Python”. My computer programming background includes several undergraduate courses, a two-week bootcamp at the General Assembly, and some experience on the job; however, I struggled mightily with two of the required classes in the program.

The Classes

Introduction for Computing for Data Analytics

Stress. Anger. Frustration. As any of my roomates at the time could tell you, I was in constant disarray during my first class in this program. Introduction for Computing for Data Analytics reminded me of my undergraduate experience. It was akin to how we described Calculus I and II for STEM majors at the University of Maryland. The weed-out classes. According to Georgia Tech’s distribution report, about 17 percent of students ended up withdrawing from this course since 2018.

The lectures for this course only brushed the surface so the majority of python programming was learned on our own. Luckily, there is an abundance of python tutorials and guidance found online. While the weekly homework assignments were fair, the exams were anything but. We had multiple, timed 2-day exams. That might sound okay, except that the average person needed 10-12 hours to complete each exam. Why do you need 10 hours to sufficiently test whether someone grasps python? That is a completely unrealistic scenario in the real world. I learned a lot from this class, but the exams left a bitter taste in my mouth.

Data and Visual Analytics

What is the optimal number of software tools to focus on during a semester? If you answered 10 or higher, then this is the class for you! This advanced core requirement combines a hotchpotch of different visualization and big data tools, throws them into four extensive projects and calls it a day. To a non-analytics individual or an executive, this might seem logical; however, technical positions require considerable depth to become experts. This is true from a concept point of view, as well as, for software tools because the concepts translate so well and BI tools are often similar. If you spend only a week learning Tableau, I guarantee you will forget almost everything just a few months later. However, if you spend an entire semester on an old-school MySQL database, many of these concepts and skills will seamlessly translate to a modern database environment in the cloud. I am seeing this first hand with a colleague of mine at work who has picked up data engineering in Azure incredibly fast because of his 10+ years of SQL database experience. This course requires an immense amount of effort, averaging 14 hours per week, per the OMS Central course review site. This means that about 50 percent of students require more than 14 hours per week. That is a tough workload for someone who has a family and/or works full time. I would argue that it is an unfair amount of time to expect for 3 credits, but I digress. Unfortunately, the poor design of this course leaves the student with an unequal balance of putting in more effort than knowledge that they leave with.

Simulation

Simulation was one of the best courses I have taken in my life. Professor Goldsman’s overwhelming enthusiasm for all things simulation can turn any math hater into a nerd. To date, he was the only professor to add some character into the pre-recorded video lectures in this program. There were constant corny jokes, links to youtube music videos, and flashing colors or noises. As any student understands, enthusiasm and creativity can make a world of a difference when it comes to learning.

While this course is more math-heavy than most in the program, the topics are incredibly useful for so many types of jobs. Simulation experiments are found in every industry, from traffic simulation to patient flow in a hospital. The class covers subjects such as probability distributions, random variate generation, poisson processes, and designing simulation experiments in ARENA. If you are interested in what these topics cover at a more granular level, I consolidated the lecture notes here.

Community

There is an old saying that it doesn’t matter what you know, it’s who you know. While it is tough to network in-person or build long lasting connections through an online program, I think the most valuable part of this experience has been the online community. The OMSA program administers a Slack network, including specific channels for each class, job postings by students, and more. Every hour of the day you will have interesting posts from thousands of intelligent students and TA’s from around the world. I posted a poll in our Slack channel to see what everyone’s strongest subject was prior to starting the program. Interestingly, the pool was incredibly diverse. About 60 percent of the students enrolled having strong backgrounds in Mathematics, Engineering, Business, or Economics. The diversity of knowledge and experience fosters incredible dialogue and a fantastic, filtered selection of thoughts, links, and notes in data analytics. It is better than any Google search!

Although the community is online, there are some opportunities to meet people in person. There are specific channels for cities with large followings of students, such as NYC, where people will schedule in person meet-ups. I attended one in NYC last year and met a few, fellow classmates.

Alternative Options and Cost

As data has exploded in popularity over the past few years, there are significantly more online masters options than I had back in 2017. However, the selection process becomes much easier once you take tuition into account. If you need a reputable program to be under $30,000, there are only a handful of options left. Here are a few programs that meet that criteria, in order of U.S. News graduate computer science or statistics rankings.

Option #1
MS Analytics, Georgia Tech
Tuition: ~$10,000 (Before fees)
CS Rank: #8

Option #2
MS Data Science, University of Texas at Austin
Starting in Spring, 2021*
Tuition: ~$10,000 (Before fees)
CS Rank: #10

Option #3
MS Applied Statistics, Penn State World Campus
Tuition: ~$27,000 (Before fees)
Stats Rank: #20 (CS Rank: #30)

Option #4
MS Data Science, Indiana University
Tuition: ~$23,400 (Before fees)
CS Rank: #55

Option #5
MS Data Analytics, Colorado State University
Tuition: ~19,000 (Before fees)
CS Rank: #75

*KD Nuggets lists additional options here.

The two alternatives to an online masters program are boot camps and self-study. I have had good experience with General Assembly, but the variability in bootcamps is huge. The importance of credentialism is changing, but I would do your due diligence if you plan on pursuing a bootcamp. As for self-study, everything I have learned in my program can easily be found online. You have a vast collection of open online classes at Coursera and MIT. My favorite resources for learning data science, analytics, and mathematics can be found here. Self-study requires a lot of discipline and time devoted towards designing a curriculum. Personally, I knew I had to enroll in a formal program in order to push myself to learn the material.

Looking Back

As I look back at my decision to enroll in the OMSA program, I can’t help but analyze whether I made the right decision. There is only so much information you can find online before you have to make a choice. Would I still have chosen this program knowing what I know now? Honestly, I don’t know, but the grass is always greener on the other side.

~ The Data Generalist
Data Science Career Advisor

Update: A part II review was published here.

If you have any additional questions on my experience with the OMSA program at Georgia Tech, please feel free to reach out. Twitter is usually the easiest.



4 Comments

Leave a Reply