Many of the techniques in conflict resolution apply to teaching, especially teaching subjects that are outside of what some people consider ordinary such as science. Of course science is and should be an ordinary thing and learning it should also be considered ordinary, but it is not in general. So I suggest that setting boundaries and establishing expectations is the first step in having a good relationship with students, of course this will be independent of subject matter but I think is especially important when teaching science.
It is common practice to require students to satisfy a minimum of competency when entering a class but in science this can be accomplished with a simple set of questions in a test. This will be the first boundary in this context. A minimum standard of knowledge for that course, both in regards to the subject and on the mathematical language required for the subject. Here is where most students fail: mathematics. The problem I think lies in the way that math is taught traditionally. Taught as an independent subject even though as examples of its implementation is common to see problems from physics like the displacement-velocity-time relationship used in differential calculus, or the simple algebra that relates mass and volume with density. But the problem arises when these mathematical concepts are applied to areas outside the traditional examples like the ones mentioned before. One area in particular that is troublesome is chemistry. It is not common to see in math courses examples from chemistry so when a student is in a chemistry class doesn't see how it can be applied. This is where establishing expectations for the course is important.
On the one hand the student must be clear on what is expected from him/her; and on the other what is the student to expect from the class, his/her peers, his/her professors and tutors, and the auxiliary material such as textbooks, and web related instruments. Addressing the first point: "what is expected from the student" is in particular difficult for several reasons. One reason of course is that students are a diverse group of people and not all have the same level of commitment or preparation. So when we think about the expectations we generalize and that might be a problem because students may feel that is not about them in particular and will not see the need to personalize the requirements established in the expectations until is too late, once the student starts to struggle in class due to the lack of preparation or commitment.
The only way I see to make the student aware of the personalization of the requirements in to do it regularly in class and not depend on the first day of class, as we normally introduce the syllabus and talk about the course in general. The teacher has to do it then but he/she has to go back to these expectations and boundaries regularly during the semester, and it has to be done intentionally and openly. What I do is to have one of the lab sessions designated to one-to-one interview where I explain clearly what I expect from them and what they can expect from me, at the same time it gives me the opportunity to learn what is their level of awareness regarding the difficulties of the course and the need for them to know and use all the resources available outside the classroom.