Seth T. Lirette is an assistant professor in the department of data science at the University of Mississippi Medical Center. He supports biomedical research across the entire university, while focusing on radiological and aging research. He is also president of the private analytics and statistical consulting firm Blackshear & Lirette. He holds an MS in statistics from Mississippi State University and a PhD in biostatistics from the University of Alabama at Birmingham.
With the data revolution firmly upon us, it has never been more imperative to have highly skilled data practitioners building models and conducting analyses. Many of those doing data science–type work have no formal training, and this often results in questionable analyses.
One potential way to rectify this situation is to provide easier access to those people who wish to pursue advanced training in analytics-related disciplines, but—for various reasons—are unwilling to take residential status at or commute to campus. While there is a litany of master’s degrees offered completely online, I am unaware of any programs for distance PhDs in statistics, biostatistics, data science, or any other analytics field. I propose it is time we loosen the reigns on the seemingly ubiquitous residential requirement for admission to PhD programs.
Disproving the sufficiency of commuting to a campus for obtaining a PhD is trivial, as we know lots of people who drive onto campuses for 4+ years and never receive doctorates. Disproving the necessity of commutation (or residentiality) is a bit more challenging. Although anecdotal, my personal experience gives a good counterexample. This essay seeks to document my experience in a distance (although officially residential) PhD program, explaining the positives and negatives alongside what worked and what needs to be improved upon.
Background
I began my stint at the University of Mississippi Medical Center (UMMC) as an unpaid intern during the summer between the two years of my master’s degree. During this internship, I was offered a full-time biostatistician position once I completed my MS degree. I accepted this position in May 2012. Then, in the spring of 2013, I was approached by who would eventually be the head of my department with an interesting proposition. He had explained that everyone was impressed with my work, and that he and the other faculty members would eventually like me to join the faculty. At the risk of sounding obvious, I would need a PhD for this to take place.
UMMC had no program related to statistics or biostatistics (we now have PhD and master’s programs in biostatistics and data science). Therefore, I needed to take the degree from somewhere afar. But I had no desire to uproot my family, nor did the department want to relinquish me. This is where the interesting proposition enters: UMMC had worked out an unofficial agreement with the department of biostatistics at the University of Alabama at Birmingham (UAB) in which I could take my PhD from UAB, but stay employed full time at UMMC. I would complete every degree requirement necessary (complete 60 hours of coursework, pass comprehensive exams, complete 24 research credit hours with a dissertation, and other requirements). The only difference between me and my student-peers would be that they were full-time graduate students living in Birmingham, Alabama, and I was a full-time biostatistician living in Jackson, Mississippi.
My Experience
From my anecdotal evidence, it seems to be somewhat common for a student to complete his or her dissertation while not living in the same locale as his or her adviser. However, I could not find a documented instance of a person doing an entire statistics PhD from a distance. So, this was going to be an exercise in faith and completely an (N=1) experiment. There were some growing pains, but for the most part, things went rather smoothly.
It was first agreed upon that I should show up for classes, at least initially. So, for the entirety of the fall semester of 2013, I made the four-hour one-way commute from Jackson to Birmingham to attend my Tuesday/Thursday classes in person and made the trek back, twice per week. This was obviously less than ideal, but everyone involved wanted to make sure I would not be falling behind on my coursework. In retrospect, this was unnecessary, as we will later see.
During the second semester, spring of 2014, I spoke with my professors, and we agreed that, given my success in the first semester, I could cut the commute back to once per week. At the end of the first year, I had logged many, many miles on I-20.
From that point on, and since my grades were exemplary, I would make it standard practice to meet face-to-face on the first day of class and then only be physically present if needed (i.e., to take an exam or to present a formal presentation).
Every professor I encountered was more than willing to be helpful in any way. Any questions I had about assignments or methods were answered over telephone calls or email exchange. I never felt as though I did not have adequate access to my instructors.
One caveat to this does need mentioning. Because this was not an official distance program, most of the classes consisted only of traditional face-to-face lectures. I took two official online classes (Writing for Research and Introduction to Epidemiology), but other than these, there was not an official platform for disseminating lectures other than face-to-face. As a result, I sometimes felt like I was not “getting my money’s worth” from some classes. I cannot fault them for this, as it was not an official distance program, but I love hearing people lecture on subjects they know intricately, and this is one thing I wish I had more of. That being said, by the time of my last course (Advanced Clinical Trials), we had a very nice virtual classroom where I was able to fully participate from four hours away via webcam. If a school chooses to fully implement a distance program, this would be vitally necessary.
Another downside to the distance model is that many people meet lifelong friends in graduate school, as I did during my master’s degree. Not living in the same city or seeing them often obviously limited the amount of interaction I had with my classmates. I knew a few through email and casual conversations, but I would have liked the chance to get to know them better.
I, however, was able to take solace in the fact that working full time provides a much better salary than that typically received through assistantship and fellowship stipends. And if I am being completely honest, my family of five needed a full-time salary more than I needed new friends.
One of the biggest advantages of working full time while pursuing the degree is that I was constantly exposed to real-world issues of data analysis. It is no secret that the curated data sets provided in classroom settings are usually nothing like data in the wild. The latter is what I was exposed to on a daily basis and would only serve to sharpen my skills. Also, with almost every new technique I learned in the classroom, I could envision somewhere at UMMC to apply it. These two worlds synergized very well. In fact, the topic of my dissertation directly resulted from one of my collaborations with the department of radiology at UMMC.
By the time it came to write my dissertation, everyone knew about my situation and I had assembled a wonderful committee willing to provide me with anything I needed to succeed.
Given my entrance into the program in August of 2013, we were initially aiming for a graduation date of May 2018. In fact, I earned my degree in August of 2017. This was, in no small part, due to both the helpfulness of the entire department of biostatistics at UAB and the generosity of what would become the department of data science at UMMC.
While I realize this option is not for everyone seeking a PhD, I hope I have made the argument that implementing a distance PhD should now be a viable option for most departments. Adding more highly trained statisticians, biostatisticians, and data scientists to the academic, governmental, social, industrial, and technological workforce can only improve some of the issues we currently face.
With more educational options, we can go a step further in adding more well-qualified practitioners of the data sciences, and we can hopefully make progress into taming what sometimes currently looks like a Wild West of analytics.
Thank you for raising this issue! I would like to add two things: my own, similar experience – I expect there are a number of good examples – and to raise an advantage of distance programs not mentioned before.
I received my PhD in statistical astrophysics, with research on time series analysis of stars and stellar populations. I maintained this program part-time while working full-time in industry sixty miles Twp years – a shorter commute with more frequent trips, usually three times a week. My world in a global industry (automotive) made me familiar with the means for distance learning and research before beginning the PhD program. I completed a traditional residency requirement and wrote a dissertation.
Two years before completion of the program, my adviser retired and moved to the Boston area, decidedly some distance from Toledo. We implemented what would now be called a Distance Learning program and she was able to remain my adviser for the completion of the dissertation.
My experience working in a global industry greatly facilitated this program. I recommend those considering distance learning consider the global environment for research today, both in industry and academia. Modern communication methods allow virtual attendance for lectures. Indeed – far from being an inherent disadvantage, making the best of a bad situation – my own experience commends distance learning: I would recommend it for most students, for at least a semester, to introduce them to how research is conducted in today’s increasingly global environment.