For the last 2+ years I've been working on an online Masters in Statistics from Texas A&M. With a May graduation date fast approaching (!!!), now seems like a good time to reflect on my experience in the program/ whether this has been a valuable use of my time. Hopefully this is a useful resource for anyone considering TAMU's program or an online degree in general.

Why an Online Masters in Statistics, and why Texas A&M?

I chose to pursue an online Masters in Statistics for a few reasons:

To improve my resume: I didn't think that I could succeed as an analytics professional in a world of Machine Learning/Big Data/AI buzzwords with only an undergraduate economics degree.
To understand what I already "knew": Undergraduate econometrics classes had certainly covered topics like regression theory, but I never felt I had a good handle on material that often seemed rushed.
To learn cool new stuff: Nate Silver had told me that Bayesian statistics was the way to go...but how did that work exactly? Time series analysis was also a major blindspot.

Why Texas A&M?

Reputation: Texas A&M has a top 20 statistics department (https://www.usnews.com/best-graduate-schools/top-science-schools/statistics-rankings) and an established online program for distance students. My roommate recommended it. Many other schools were just ramping up their online offerings at the time of my application, and being a guinea pig didn't seem like the best idea.
Cost: A Masters degree for $35K felt like a steal (and it could be even less for you if your employer can be convinced to pay for some of it!)
Course selection: Texas A&M offers a strong core curriculum supplemented with interesting elective options (http://online.stat.tamu.edu/course-list/). I appreciated that there was clarity on when each class would be offered- I also never had any issues getting into a course.
Football: For all his flaws, Johnny Manziel was the most exciting college football player of his generation. Who wouldn't want to root for the Aggies?

Distance Student Logistics

Almost all courses in the program are taught live to on-campus statistics students while being simultaneously offered to distance students online.

Distance students do not listen in live to class. Instead, lectures are recorded and uploaded online (usually within a few hours) for us to download and view later. Compared to the undergrad experience of dragging myself from dorm to classroom each day, recorded lectures were glorious. I could pause and rewind at any time, or increase playback speed to 1.4X or 1.6X if the material was easy. I could watch lectures in the evening after work, or stack a week's worth together on a Sunday afternoon.

Exams are taken in-class for on-campus students, while distance students generally have a 24 hour window to take exams under proctor supervision. A proctor is someone tasked with administering the exam and basically making sure you don't cheat. You can register someone you know to act as a proctor (with a few restrictions) or else go to a testing center, usually at a local library or university.

Texas A&M uses eCampus software to administer online classes. It works surprisingly well. Professors can post assignments, code, past tests, and announcements easily, while there's also a discussion board for students to ask and answer questions about homeworks or exams. Most classes also have a weekly live Q&A session run by either a TA or the professor for additional help. These sessions came in handy for covering tricky HW problems or going over review questions prior to exams.

Course Schedule

36 credit hours are required to graduate. My schedule was as follows:

Spring 2015
STAT 630: Mathematical Statistics
STAT 604: R/SAS Introduction
Summer 2015
STAT 608: Regression
STAT 681: Seminar (1 credit hour)
Fall 2015
STAT 641: Methods of Statistics I
STAT 607: Sampling
Spring 2016
STAT 642: Methods of Statistics II
STAT 657: Advanced SAS Programming
Summer 2016
STAT 626: Time Series
Fall 2016
STAT 638: Bayesian Methods
STAT 684: Consulting
STAT 685: Directed Studies (a research project that stretches over two semesters)
Spring 2017
STAT 685: Pt 2
STAT 659: Categorical Data Analysis

STAT 630, 641, and 642 were the most time-intensive courses. I recommend taking them individually or pairing them with a manageable elective.

The Dreaded Qualifying Exam

It turns out that you don't just have to complete 36 credit hours to graduate. You also have to pass a qualifying exam covering the "core" material (STAT 608, STAT 641, STAT 642). The exam is only offered twice a year (January/August). I took it in August 2016.

The exam is four hours long. You don't get a formula sheet. Fomulas must instead be memorized like in a Victorian schoolroom. Studying was miserable. I recommend giving yourself at least 4-5 weeks of lead time.

On the plus side, you get access to literally decades worth of prior qualifying exams to aid preparation. It's only pass/fail. And passing gave me a legitimate sense of accomplishment.

Looking Back, The Good Stuff:

The program has given me a strong grounding in distribution theory, regression theory, basic time series and Bayesian analysis, sampling theory, and other topics. I'm nowhere near an expert statistician, but I know a ton more than I did two years ago.
In general the quality of teaching at Texas A&M is excellent. Professors are extremely attentive to answering distance student questions, either through the eCampus discussion board or at Q&A. The administrative staff also does a great job informing distance students of deadlines that might have been missed othewise.
Having on-campus students in the same class was very helpful, as often a question would be asked during a lecture that would clarify the topic at hand.
Aggies' constant use of "Howdy" has really grown on me.
The proctoring system works relatively well for exams. I think it strikes the right balance of deterring cheating without being completely onerous.

Looking Back, The Not So Good Stuff:

Many TAMU professors are old school. This means they use SAS. A lot. As someone falling squarely in the R/Python generation (see relevant kdnuggets chart below), parsing SAS code for various assignments did not feel like a productive use of time.

http://www.kdnuggets.com/2016/07/burtchworks-sas-r-python-analytics-pros-prefer.html

Some of the methods taught are classical to the point of being outdated. Discussions on variable selection for example usually focused on forwards/backwards/stepwise selection instead of regularization methods like the LASSO. "Big Data" is not really on the table here.
Having several distance students (all working full-time) coordinate on a group project, as one elective class did, felt like cruel and unusual punishment.
Many of the business questions I'm coming across now deal with ascertaining causality from observational data. Unfortunately causal inference methods have not been a part of any of the classes I've taken.

Final Thoughts:

Overall I think A&M's Online Masters in Statistics program has been excellent value for money. While some of the course material could use some updating, the core statistical theory taught is as applicable now as ever.

In addition, I will soon be an alumni of an SEC powerhouse with some delightfully weird traditions. GIG 'EM.