Many years ago, my univerity's choice of calculus book also had the solutions to half the problems (evens or odds, doesn't really matter).
Department policy was also one of "we won't provide you the correct answers" even after an assignment had been turned in because the book was used for multiple years. The publisher had a new edition every year, but the university stuck with the same book because it was used for Calculus I, II, and III, which for most students was 3 or 4 semesters between starting I and finishing III, usually due to a scheduling conflict requiring a semester off between them, or because they had to repeat Calc II since the math department's selection of instructors was particularly bad for that course.
Those same (usually bad) instructors were all too happy to follow the department policy and not provide any feedback other than "correct" or "incorrect".
On the other hand, the university book store carried, and put on the shelf right next to the calculus book, the publisher's "teacher's solution guide", in two very reasonably priced volumes, which had the answers for the other half of the problem sets, as well as the step-by-step process for most of them, which was the valuable part, as you could see where you were erring.
Math department policy was that you weren't allowed to have those, either.
You can guess how well that policy was followed.
Those that put the effort in and learned the material did well. Those that just copied from the solutions book and turned it in did not.
Some students are very good at taking exams, others, not so much. This can be due to a learning disability, stress over test taking, etc (or it could even be that person just is having a bad day!). Having homework and projects allows for students to have a different way of showing that they understand the coursework, and are able to apply the material.
It also gives the professor and TAs insight into the student. Why is a student doing so well on homeworks, and not the exam or project?
Some courses have it where if youre final grade on the exam is an A, you get an A (since it is a comphrensive knowledge base test of what you are expected to know of the material). But, let's say you don't do as well on the exam, you can have the homeworks average out the exam grade. Or you have a project, that can help equalize out the grades, because the application of the knowledge is important as well.
If the person teaching a maths course needs to look the answers up, you are just fucked.
We have a set of homework assignments to help the students learn the tools (kind of a "hello world" script for the project), and I asked if we just tell them we don't look at them and they are entirely for your benefit. The professor said you if you do that, there will be a decent amount of students who will just not do the assignments and later on complain that they have no idea how to use the tools.
The solution we came up with is to have a set assignment due date so students feel like they have to do them.
Personally, I am surprised that graduate students need that sort of motivation to do the work.
And really, shouldn’t this be how we approach most problems?
You're right, of course, but then it seems like the question is the following: which of the two should the class serve?
I know, I was referring to the homework.
And then give the answers after the homework was handed in, if they want to grade the homework. Or compromise, by giving answers to half the assignments, and not grading those.
That being said, I seem to recall that when I took math classes in college, we had books that had even questions that the solutions were in the back, and the odd questions did not (they were only provided to the teachers). The questions were largely the same, so if a student had an issue with an odd question, they were able to just go to the cooresponding even question and work it out. I felt that was a reasonable compromise.