Yes, they were sort of helpful except he never gave straight answers to my questions. I think that's a philosophy thing though.
More seriously, I've identified a few things I need to work on, and a few things about the format of the course that caused serious problems.
First, as a grader, I'm used to working through math exams, not papers. Grading math exams, you usually only need to identify the point at which the student forgot how to use the method she or he is applying in each particular problem. Paper grading is much more subtle, and I need to take the time to figure out the most time-efficient way to make helpful comments.
Second, I was surprised to see several students say things like `He also never really encouraged us or made us feel positive.' I need to give more positive feedback, even just for answers that are on the right track but not totally right, in both class and papers.
Third, especially with introductory courses, the students need a fair amount of explicit structural guidelines when they go to write their papers. A common strategy I've seen my friends and colleagues use is to have a series of progressively more involved paper assignments. For example, in the first assignment, students just have to summarise an argument in a perspicacious way; in the second, they summarise and formulate a single objection; in the third, they summarise, object, and then rebut the objection. The distribution of the paper assignments is accompanied with discussions on how to summarise arguments and how to formulate objections.
Fourth, students need some variety. Lecture-lecture-discussion gets old, and sometimes they need more lecture or more discussion. That's a flexibility I don't really have as a TA, but I do have some wiggle room: I can sacrifice our discussion time to lecture, and do things like ask students to give presentations or lead discussions.
Fifth, I have learned how not to formulate exam questions. Intro students should be asked to regurgitate major arguments, or briefly compare and contrast what Plato, Descartes, and Perry think about the soul. They should not be asked to recall trivial minutiae about the readings, with virtually no guidelines as to what minutiae they will be asked about.
And sixth, at the introductory level, a `student directed' course is just begging for disaster. I'm not sure that I ever explained the format here, so let me do so briefly now.
Thursday night, the students would finish their assigned reading, and email questions on that reading to their TAs. We sorted through them and put together handouts, and the discussion section on Friday was usually spent working through these questions. We would usually get through 4-6 (out of 20+ or so distinct questions) in the fifty-minute section. At the end of section, the students would identify questions that the TAs would send on to the professor that night. The lecture periods on Monday and Wednesday consisted almost exclusively of the professor answering these questions. Hence, if the students didn't consider it important or interesting or confusing, or important or interesting or confusing enough to warrant being sent up to the professor, it was never discussed in the context of the class. To make matters worse, we were supposed to never give them straight out answers; they were supposed to struggle with answering the questions of their classmates, and get direction but not answers from the lectures.
Unsurprisingly, many students found this immensely frustrating and never achieved even the barest beginnings of understanding of the texts. Incorporating at least a little Textbook Philosophy would have been slightly frustrating for us the professional philosophers, but at least the students have a chance of understanding what's going on.