Becoming an External Ph.D. Student

Within my short time of being a professor at a German engineering university, I’ve raked up quite a number of requests from folks to become an “external Ph.D. student” of mine. In the German system, this is someone who is enrolled in the Ph.D. program but not funded by the university, i.e. is self-sponsored. Since you are only allowed into the Ph.D. program after a successful (and appropriate) Master’s degree, anyone with that degree is ABD, “all but dissertation”. So folks approach me with a research idea, typically about some itch they are having as a developer, and ask whether this would be enough for a dissertation and then the title. My answer, naturally, is “it depends”.


In theory, according to the Ph.D. program rules, someone could slam a dissertation on my table, ask to be enrolled in the program, and graduate on the (assumed) strength of that dissertation within a couple of weeks. That is, if I’m willing to accept that person as my Ph.D. student. And given that there is no or only token tuition in Germany, there is little downside for me to reject a person if the work is good. Looking at FAU’s Ph.D. program rules, however, you can read that it is strongly advised that the external Ph.D. student engages with me early so that I can guide the research. The goal is quality assurance through process, I’m guessing, even though that does not speak against a genius finding his or her way on their own. With that, it implies I have to invest time.


I haven’t had an external Ph.D. student graduate yet, but I have had many folks who were very enthusiastic about their research idea, engaged with me high-energy, only to be never heard of again. So this is a problem of persistence. Doing Ph.D. level work on the side, while you are working in a regular job, is very hard. You need to be seriously dedicated. I can make the safe assumption that most people won’t have sufficient energy and persistence. But it is hard for me to know beforehand, and in particular, I can’t derive a test from that to see whether my time engagement with a requestor makes sense or not.

What I can check is their initial qualifications. Most folks are smart enough to engage early and not to show up with a fully-done dissertation. They ask for advice, whether their idea makes sense, and whether I would be willing to be their advisor. From my short experience, the biggest stumbling block that I can measure is the requestor’s research education. Sadly, most students never received an education on how to perform scientific work. They’ve learned programming, architecture, algorithms, and processes, but for work in industry. They never learned to do interview analysis, or set up experiments and carry them out, or properly validate their results.

A dissertation needs to advance the state of the art and has to show this convincingly. If someone was never taught appropriate research methods, he or she is unlikely to pick this up on the side. They need an advisor to guide them through the process of learning of how to properly frame a research question, choose approriate research methods, execute those methods well, and present the results in a digestable manner. If someone hasn’t learned how to do this, and they still want to write a dissertation, they need to learn how to do research first.

For people local to the Nuremberg region, I recommend my research lab course Nailing your Thesis as a starting point.