It is well known that professors and undergraduates exist on different planets with respect to their expectations and views about educational issues (like grades). That may relate to the difference in their ages, or in the intensity of their academic focus. Those factors are less pronounced in the relationships between professors and graduate students, who, nonetheless, also exist on different planets and have different views about the graduate-research experience.
For example, some graduate students, including research assistants, believe that they are exploited, employed at low wages to work long hours accomplishing various tasks that benefit the research endeavors of an adviser who doesn't really care about them and whose own "work" may not be apparent to the student. I don't doubt that there are cases in which that description applies to a particular professor, but it's not an accurate description of the typical graduate experience, at least not in the physical sciences with which I am familiar. It's an incomplete and inaccurate description for at least three reasons.
1) Not cheap labor. Graduate-student stipends may be low compared with other employment options, particularly in science and engineering fields, but students are not "cheap" labor for advisers. When salary, benefits, and, in some cases, tuition are factored in, graduate students cost a lot, and most or all of that cost may come out of the adviser's research grants.
Graduate students don't see those additional costs; they just see their modest salaries. In fact, graduate-student salaries and their related costs may largely consume grants. Expenses for the actual research may be the smallest component of the budget.
From the adviser's point of view, therefore, students are getting paid a decent (living) wage while working toward their (tuition-free) graduate degree, and doing interesting research in the process. A student, however, may focus on how hard the work is for not a lot of money in a stressful environment that may be populated by some intense and/or difficult people. If the student has or wants to have a family, the stipend may seem even smaller. Financial pressures may be a source of discontent on both sides because each has a different perspective on the "cost" of the research.
2) Training time. Most students do not arrive in graduate school knowing how to do research. It takes time to learn. Unlike most postdocs (who have already successfully attained a Ph.D.), some graduate students never learn.
If the training time and the uncertainty that a graduate student will do well in research are factored in, one could reasonably conclude that using students is an extremely inefficient way for an adviser to conduct a research program. A student may need time to adjust to a new environment in which expectations and skills are different from those in a typical undergraduate program. At first, the student may be taken aback by the culture of criticism, discussion, and debate of graduate seminars, research-group meetings, and research presentations.
Some students can handle all of that and some can't —no matter how smart they are. In fact, from the professor's point of view, the most efficient way to conduct a research program would be to hire nonstudent workers who are already trained and who would stay in the job on a long-term basis rather than leaving just at the point when they finally know what they are doing. That would be more efficient even than hiring postdocs who only stay a couple of years and then move on.
That would be fine if efficiency were the only thing that mattered, but a completely efficient scenario of trained workers doesn't sound appealing to me, nor does working in isolation. Most of us science professors aren't here to manage a group of technicians, or even to work alone.
I do like to get results, and my fondest wish is that students who are paid on a grant will get some results, for their sake and mine. But I also expect a bit of inefficiency along the way. By results, I mean data, a talk, a paper, or a new grant proposal. We need such results to keep the interconnected system of research and graduate education functioning. Advisers may be more focused on certain important deadlines (including those involving tenure and promotion decisions) than students and may transmit (without much explanation) their stress and sense of urgency to their students.
In that context, the concept of efficiency doesn't capture the most valuable outcomes of teaching students how to do research, whether the teaching involves direct instruction or letting a student loose on a problem. The most valuable outcomes are discovery, insight, and inspiration (and having fun in the process). Can those be taught? Years of advising lead me to an unsatisfying answer: Sometimes.
3) The way we work. Most students, even quite senior graduate students, have little idea of what faculty members do all day. I have heard students complain that they do all the work while their advisers do nothing. I am always skeptical that a professor managing a research group at a research university is really doing nothing all day.
There are some periods of time, including entire academic terms, when I don't have time to do any actual research myself. I suppose in some respects I am doing nothing during those times —nothing other than teaching, serving on committees, reviewing manuscripts and proposals, writing manuscripts and proposals (an activity I count as research), dealing with budgets and accountants involved in grants management, writing letters of recommendation, attending conferences (preparing and giving talks), and a host of other random things that seem to pop up every day and consume my time.
When doctoral students graduate and become faculty members, perhaps after doing postdoctoral research, a common refrain is "I didn't know I would have to spend so much time doing … [fill in blank with administrative or advising task]."
We advisers could do a better job of teaching our students exactly what professors really do. That might result in less dissatisfaction at a perceived imbalance in workload between students and their advisers. Students should also be more aware of the environment in which they are working, although some of what is involved in being a professor and adviser of a research group is difficult to anticipate or understand until you actually do it.
I like having a research group, and I like working with students. I enjoy doing research, discovering things, developing new ideas, and communicating the results, and I like trying to teach others how to do all of that as well. It takes a lot of time and energy for both adviser and student, even when things go well and even when the student thinks he or she is doing most of the work.
Some advisers are more involved with their students' research and education than others. Some leave a lot of the day-to-day advising to other members of a research group. Some advisers would prefer to have more "workers" and fewer students, especially advisers who have had a lot of negative experiences with unproductive graduate students. It can be extremely frustrating and demoralizing to (try to) work with a dysfunctional grad student.
I think, however, that most of us advisers have enough positive experiences to balance out the negative ones —even if the negative ones are rather spectacular and make for better stories.
By working with many different students over the years, we can achieve a reasonably upbeat perspective on the overall experience. In contrast, most graduate students work with only one or two advisers, so a single bad experience can be crushing.
Most of us science-professor types at research universities advise graduate students, for better or worse. Sometimes it works out and sometimes it doesn't. Successful adviser-student interactions require of both parties a balance between being patient and being assertive, keeping overt complaining to a minimum, and realizing that what seems like insensitive or strange behavior or laziness in the other might have a reasonable explanation.
Graduate students and professors alike are continually amazed at each other's mystifying behavior, so it is not surprising that there are gaps in experiences and expectations between them. But maybe it's not surprising that these misunderstandings exist: My colleagues and I often don't understand each other, either.