4 ms·

The best class I've ever taken was an intro class in statistics, in which the professor recommended... no book. He had an entire set of notes, practice problems

The best class I've ever taken was an intro class in statistics, in which the professor recommended... no book. He had an entire set of notes, practice problems (with extraordinarily detailed answers), and problem sets. Every year, he creates new exam problems so the old exams with answers become practice material. Honestly, his lecture notes (and all the additional material) is so vast that a student would hardly have enough time to study anything outside the notes. My point is, if a lecturer is really dedicated to providing the best possible experience to his students, he doesn't need a textbook. I'm not saying that it's easy, or that it doesn't require a lot of time and effort from his part. It does. Some lecturers chose to accept the load, while others go the easy way and just point to references and exercises in the book.

_delirium 14y ago
That answer is essentially, "each professor should write their own textbook", since an extensive series of notes intended to be used in a course of study, along with practice problems, is pretty much a draft textbook. That could be an interesting solution, but is mostly practical if you already have tenure, since it is a quick way to not get tenure otherwise. If you spend your junior professorship time writing a textbook (in a field that already has many) rather than spending that time on more research and publications, tenure is unlikely. If you're tenured and have been teaching the same course for a while, on the other hand, you both have more leeway to spend your time doing that, and may have already sort of "naturally" accumulated years worth of notes from previous iterations of the course. Then it can be a good idea, and many senior professors do prefer this approach, because it lets them teach the course in a way precisely in line with how they feel the material should be taught, rather than having to buy into someone else's approach. Universities could influence that outcome at all levels if they wished, though. Universities formally budget preparation time per course. Currently it's typical to budget quite little, because the assumption (and sometimes requirement) is that you'll be teaching using a standard text in the field. If they instead want extensive preparation to be done, along the lines of what you propose, they can: 1) budget sufficient time for it, e.g. specify that each course requires, say, 100 (or 200, or whatever) hours of preparation prior to the beginning of the semester; and 2) expect that the preparation time be used towards the budgeted end (i.e. expect extensive independent course materials to be developed for each course, in line with the amount of time budgeted).
- dhimes 14y ago
  Furthermore, the first edition of most[1] textbooks suck- draft books much more so. I think that this would be a recipe for disaster. [1] Note to the pedantic: I am taking license with the fact that I've only read a few first edition textbooks- and they sucked. "Most" in this case should be read as a rhetorical vehicle.
- seanmcdirmid 14y ago
  > "each professor should write their own textbook", since an extensive series of notes intended to be used in a course of study, along with practice problems, is pretty much a draft textbook. This isn't really true. A majority of my math text books at UW were written in note form by the professor teaching the class. They were just very efficient...and hey, its math where a glossy text book really gives you nothing. > Currently it's typical to budget quite little, because the assumption (and sometimes requirement) is that you'll be teaching using a standard text in the field. This is definitely not true in math and barely true in computer science. I think it is true in physics and chemistry, but only for entry-level courses. It also depends on the university. Open sourcing material between professors could go a long way; each could simply collect what they want for the course and then easily publish it in electronic (free) or paper form (students pay printing costs). I think this is how it mostly works already in upper division courses. Sometimes there is a very nice up to date book; but just as often, say in computer science, everything is out of date anyways (computer architecture, compilers, ...).
  - _delirium 14y ago
    > This is definitely not true in math and barely true in computer science. I haven't surveyed everywhere in CS, but I know the operations of quite a number of schools, and I don't know any that budget significant preparation time for "normal" classes. Some do allocate extraordinary preparation time if you're developing a completely new course that's never been taught at the school before. But in the common case, where it's a longstanding standard course, you're expected to use reasonably standard material, whether it's whatever was used in the previous professor's version of the course, or some moral equivalent that you're expected to be able to adopt within <10 hrs. You definitely aren't supposed to (officially) spend more than a minor amount of time preparing your own curriculum and pedagogical materials for the course.

_delirium 14y ago

That answer is essentially, "each professor should write their own textbook", since an extensive series of notes intended to be used in a course of study, along with practice problems, is pretty much a draft textbook. That could be an interesting solution, but is mostly practical if you already have tenure, since it is a quick way to not get tenure otherwise. If you spend your junior professorship time writing a textbook (in a field that already has many) rather than spending that time on more research and publications, tenure is unlikely. If you're tenured and have been teaching the same course for a while, on the other hand, you both have more leeway to spend your time doing that, and may have already sort of "naturally" accumulated years worth of notes from previous iterations of the course. Then it can be a good idea, and many senior professors do prefer this approach, because it lets them teach the course in a way precisely in line with how they feel the material should be taught, rather than having to buy into someone else's approach. Universities could influence that outcome at all levels if they wished, though. Universities formally budget preparation time per course. Currently it's typical to budget quite little, because the assumption (and sometimes requirement) is that you'll be teaching using a standard text in the field. If they instead want extensive preparation to be done, along the lines of what you propose, they can: 1) budget sufficient time for it, e.g. specify that each course requires, say, 100 (or 200, or whatever) hours of preparation prior to the beginning of the semester; and 2) expect that the preparation time be used towards the budgeted end (i.e. expect extensive independent course materials to be developed for each course, in line with the amount of time budgeted).

dhimes 14y ago
Furthermore, the first edition of most[1] textbooks suck- draft books much more so. I think that this would be a recipe for disaster. [1] Note to the pedantic: I am taking license with the fact that I've only read a few first edition textbooks- and they sucked. "Most" in this case should be read as a rhetorical vehicle.
seanmcdirmid 14y ago
> "each professor should write their own textbook", since an extensive series of notes intended to be used in a course of study, along with practice problems, is pretty much a draft textbook. This isn't really true. A majority of my math text books at UW were written in note form by the professor teaching the class. They were just very efficient...and hey, its math where a glossy text book really gives you nothing. > Currently it's typical to budget quite little, because the assumption (and sometimes requirement) is that you'll be teaching using a standard text in the field. This is definitely not true in math and barely true in computer science. I think it is true in physics and chemistry, but only for entry-level courses. It also depends on the university. Open sourcing material between professors could go a long way; each could simply collect what they want for the course and then easily publish it in electronic (free) or paper form (students pay printing costs). I think this is how it mostly works already in upper division courses. Sometimes there is a very nice up to date book; but just as often, say in computer science, everything is out of date anyways (computer architecture, compilers, ...).
- _delirium 14y ago
  > This is definitely not true in math and barely true in computer science. I haven't surveyed everywhere in CS, but I know the operations of quite a number of schools, and I don't know any that budget significant preparation time for "normal" classes. Some do allocate extraordinary preparation time if you're developing a completely new course that's never been taught at the school before. But in the common case, where it's a longstanding standard course, you're expected to use reasonably standard material, whether it's whatever was used in the previous professor's version of the course, or some moral equivalent that you're expected to be able to adopt within <10 hrs. You definitely aren't supposed to (officially) spend more than a minor amount of time preparing your own curriculum and pedagogical materials for the course.