During my PhD at TU Delft, I had the opportunity to supervise 3 MSc thesis students and a teaching assistant. I've come a long way from dumping all my knowledge and all the papers on poor young students, and have been striving to find a balance between guiding students and giving them enough space and time to let their own thoughts blossom.
If you have the opportunity to supervise MSc thesis students during your PhD, I would strongly recommend your doing so. Not only can they help on some peripheral research questions, ut they also make you grow as an academic. Once you graduate with your PhD, you are expected to be an independent researcher, and someone who can take the lead in research projects that may employ researchers in more junior stages of their careers.
Here are the 7 main lessons I learned from supervising students during my PhD:
The first challenge lies in identifying a good MSc thesis topic. You want the student to work on research related to your own, so you can for example look at a smaller open question that remains in your work. You could as well see if an experimental technique you are using can be applied to study a different type of sample. Or you can get delegate the coding of a tool that you might need.
Keep in mind that an MSc thesis topic is significantly smaller and with less theoretical depth than a PhD thesis topic - so make sure you find a piece of work someone can successfully finish in the given amount of time (typically 1 year).
The thesis topics of my MSc students dealt in 2 cases with the nonlinear finite element modeling of my experiments (something we had defined as not being part of my research), and giving me a hand in the lab. Another student started off from nonlinear finite element models to study in more detail the interaction between punching and shear, by running a number of computer tests that were benchmarked with my experiments.
Taking on students who will work with you, implies that you need to be able to communicate the gist of your research and teach your most important findings to them, so that they can build on your work. Explaining all the details and procedures of your research will not be necessary, but to find the most important parts of your work and communicate them, is a skill that will serve you well at conferences and to communicate with a broader public about your work.
What is the most important part of your research that the student really needs to understand? Maybe not everything you calculated, tested or modeled is equally successful or important. Take a step back from your work and honestly evaluate the picture you see in front of you: what are the most important outlines of your work that you need to communicate to someone who wants to work on this topic? Again, this skill will help you communicate your research.
What are the key references a student should read? What are the basic assumptions of your work? When you take on people who work on the same topic as you, you need to be able to tell them about the fundaments of your work: what is your work based on? What is essential to your research? Avoid pulling in all the interesting information and bombarding students with every possible interesting paper - go back to the clean and pure starting points of your work, and share this basis with your students.
The joy of working with students lies in that moment when they start to come up with their own ideas. When they come and propose something you hadn't thought of, when they have a working model, when their approach suddenly fits yours, when they have an interesting results in the lab,... those moments are the culmination of your work as a supervisor. I clearly remember those moments of joy with all students I supervised - it is indeed a special feeling, and something you will see repeated in the future as you work with more students at different levels. Give them time and space so that their work can bear the fruit and result in new ideas.
Even though you might feel tempted to start explaining a student in detail how you think he/she should tackle his/her research question step-by-step, hold your horses. Don't overload students. Don't spoonfeed students. Just give them some basic information, and then send them out on their own. Doing so, you will leave more room for their own creativity and research abilities - which, in term, might result in them surprising you with new ideas.
All the previous steps come together in learning how to be, what I call, a "gentle guide". Be gentle, don't force your thoughts upon students, and leave them space to explore new options. At the same time, you need to be a guide: give them some guidance, introduce them to the world of research, give them their little starter toolkit, and be there when they need advice (about the research, or when they freak out and think they'll never graduate in time). It's a process, and it also means being able to let go of some parts of the research that you might have considered as completely yours, and it means opening up to new research talent. And that, my dear readers, is in a nutshell what you will be doing as an independent scholar as well.