Is an online degree worth it?
Several months ago, I graduated from the University of Illinois Urbana-Champaign, Master of Computer Science.
Note: I have a newer and more comprehensive review (with subject details) here:
My background was in statistics and economics, and programming was something I learnt on the job. Needless to say, I lacked the rigour of a computer science major, and this was something I wanted to patch up.
I enrolled in early 2018 where there were only a few legitimate options available for someone that wasn’t based in the US. Now, I can see a explosion of alternatives. Many of them run off the Coursera or edX platform for content delivery. Here’s a couple…
OMSCS @ GATech http://www.omscs.gatech.edu/explore-oms-cs
MCIT @ UPenn https://www.coursera.org/degrees/mcit-penn
MIDS @ UCBerkeley https://datascience.berkeley.edu
The UC Berkeley degree is offered through the School of Information, and therefore more focused on data science. This along with the MS @ Columbia University are extremely expensive at 50k+ and therefore out of my budget. The rest are around 10–20k USD — very reasonably priced for a Master degree. At the time the only two affordable offerings that were available to me were UIUC and GATech. They are both extremely reputable universities. (http://csrankings.org)
The influx of these degrees suggest two things:
- There’s increased demand out there, in part due to the data science / AI / machine learning boom. This is augmented by the fact that online learning is no longer taboo. Certainly with Covid19, it is almost becoming the norm!
- It seems as though a lot of universities are really struggling for additional revenue. These master degree courses, which are basically a mish-mash of ugrad subjects re-shuffled somewhat, have become a cashcow for universities struggling to get sufficient government funding.
With that out of the way, here’s an overview of my experiences at UIUC.
Most of the students are based in the US, I would say circa 60–70%. Your average student would be in their mid 30s to early 40s, most likely with spouse, kids, and a day job. This was nice as it meant many already had significant hands on experience. Due to other commitments, most students do one or two subjects per semester.
UIUC’s MCS course material was delivered through Coursera. Exams were conducted through ProcterU. Formal student / tutor interaction was conducted through piazza and informal student chit-chat was done through slack. The student community on slack was extremely friendly, helpful and collegiate.
Most subjects had between 1 to 2 exams and 2 to 3 major projects. For more math heavy subjects, weekly homework assignments were given. These usually took at least 3 to 5 hours to complete — and sometimes much longer.
Aside from tuition fees, approximately 20k in total, you have to pay Coursera fees which were circa 100–200 per subject and ProcterU exams — another 10–20 dollars per exam.
I really appreciated the flexibility. If you undertook a degree at your local university, you would be in trouble if you wanted to move interstate or overseas for your job. Online degrees — you can do anywhere. Though I did hear some issues about Chinese students and their VPN connections, especially when doing ProcterU. So if you need to be in China for a prolonged period of time, you might want to look into that further.
I found the overall difficulty of the courses to be lower than expected. Marking was generous, especially for group assignments. This is somewhat to be expected, as you had a small handful of tutors marking a large cohort of online students. Harsh marking would only lead to student complaints and more emails and more work … At least that was the feeling I got. Often assignment marks were released very late into the course and sometimes with very little feedback. However, exam difficulty was largely fair and in-line with my expectations.
What I notice was that the quality of courses varied quite significantly. As a general rule, I found the machine-learning and statistics courses were better designed and more rigorous than the computer science courses. This was unfortunate for me, because I already had a statistics major, and really wanted to build some rigour on the computer science side of things. An exception to this rule was CS 410 Text Information Systems, which was put together brilliantly — I would highly recommend it.
Enrolling subjects online was a painful experience. Sought after courses would be fill very quickly (within an hour or two of openning). If you live in the wrong time zone it can be very problematic.
An online master degree provides you with some breadth and some employer (or future employer) recognition. If you’re transitioning from a different field into tech, it makes perfect sense. If you’re looking for robust and rigorous theory and programming, it seems to be lacking in both. This unfortunately is something you need to tackle on your own (perhaps armed with new enthusiasm and a new degree).
Create your own data science degree online: