“In applied sciences like software engineering, PhD students should not work at the university full time.”
This proposition relates somewhat to Proposition #5. I think it is very important for researchers in software engineering to connect with practice. Doing industrial case studies is one option, but I want to go even further, I think young researchers should be prepared for a life a real world researcher by spending time outside of the university.
During my PhD, for 1.5 years, I spent one day a week at a Robeco, a large Dutch bank. This happened due to a great amount of coincidence: at a developers meeting, I happened to be standing in the line for coffee next to the head of the end-user computing team of this bank, but it changed the entire course of my dissertation. It helped me in the following ways:
I have a lot about PhD students losing their motivation, some say motivation is more important than IQ for learning and as many as 37% of PhD students drop out. While, as everyone, I had some moments in the past 4 years that I considered quitting, this was never due to lack of motivation. Every Tuesday I spent at Robeco I left convinced that people were going to benefit from research.
Explaining your research
Being within a company setting, doing research is not the default thing, so over the course of these 1,5 years, many people asked me what I was doing there and that greatly shaped my thoughts on what I was doing there and trained me for explaining it quickly and simply.
Building a network
If you are going to do case studies, you will need access to and connections with companies. Spending just one day a week at a company will help you build a network with people who know you as a researcher and might help you to find new companies.
So, PhD students, try to find a company where you can spend just a few days to begin with. Engage with people, perform formal interview, work with them in their regular process. It will make your PhD project more real and more fun.