Actually, I don't differentiate whether I supervise BA or MA theses or term papers -- there are only gradual differences with respect to the amount of time to invest, the number of words or pages to produce, and the kind of topic. At all levels, I expect students to come up with a proposal first; they should define themselves which topic they work on for the next few weeks (or months in case of MA theses). We discuss the proposal and then they have to write a one-page abstract defining their research question, what kind of material to use or to produce, and they also should list first literature references. I don't expect them to read 24/7, but they should investigate relevant literature for their topic. Then we discuss this abstract and refine the research question. I also point them to unmentioned papers or journals and book series they should consider consulting. And then I send them off to work.
Usually, there is a fixed date for handing in the thesis or the seminar paper. When half of the time is over, I invite the students for another meeting. A few days before, they send me everything they have produced so far. At this point in the process, it's time to have a look at the structure of the text, even if some sections are still to be written. The list of references they want to cite should be nearly complete. This appointment forces them to actually write something and not spend time procrastinating. There is a strong correlation: Students with little and/or poor text half way through their paper, usually hand in poor final versions. They underestimate the time needed for revising, editing, and printing (university regulations often require printed theses).
Sometimes, students see those two appointments (preparing a proposal and discussing it, and sending mid-way material and discussing it) as some kind of cruelty. However, it helps them a lot in focusing on their topic, developing critical thinking, and producing high-quality texts. After grading their work, students can also make an appointment to get detailed feedback.
When I teach seminars where several students write a paper, I often make some of these appointments a group activity: students report on their research question in the first session, we discuss how to define a good hypothesis and how and where to find relevant literature. In the half-time session, students report on their progress and the group comments on whether to refine (expand or narrow) the research question or their ambitions. That's also a perfect opportunity to realize that all the others also struggle a bit, didn't manage to write as much text as they expected, etc.
Interestingly, the Department of Linguistics at the University of Konstanz recently developed a checklist for BA students: they should reflect on the competencies required to write a thesis (including writing competencies) and they should prepare a one-page abstract before looking for a supervisor. The checklist is intended to help students define a topic and become aware of all the skills involved when writing a thesis, and the filled-out list should inform potential supervisors about the students. As for me, I don't need this list, its elements are part of my supervision style anyway. However, preparing such a checklist is a sign for aiming at a somewhat more homogeneous supervision for all students, but you could also see it as an indication that not all potential supervisors have a real interest in mentoring students.
What's your supervision style?