There are several factors that can make a Ph.D. advisor more effective. Of course, thorough knowledge of the subject is a must, along with sufficient time and willingness to help a student. However, this is just scratching the surface of what makes such a role truly constructive. Let's drill down into the specific qualities of a Ph.D. advisor and find out what should be the ultimate goals of such an endeavor.
1. Set clear expectations
Expectations for graduate students vary widely. Different departments, professors, universities, and fields of study have vastly different expectations of their students. While some requirements may be codified by the department (e.g. each Ph.D. student must serve as a teaching assistant for a total two semesters), the reality may be quite different (e.g. most Ph.D. students teach nearly every semester because it is their primary means of financial support). Other expectations are not formally codified, but essentially non-negotiable (e.g. Ph.D. students will engage in research activities through the majority of the summer break).
To prevent misunderstandings and to set each graduate student on the road to success, clearly tell your students what you expect of them. Explain your longer-term goals for them: e.g. they should develop a reliable system to measure XYZ by the end of year 2, present new results at conferences every year, and be done with the majority of their data collection and analysis by the end of year 4 or 5. Also clarify your more mundane expectations: e.g. graduate students are expected to work a minimum of 40 hours per week, come in on nights and weekends as necessary, and take 3-4 weeks off for vacations and holidays per year. To keep students on track in the short term, tell them what you expect by your next meeting: e.g. find and read papers on topic C, do a test run for experiment B, and analyze your results from experiment A.
Obviously, these goals will need to be adjusted as circumstances change. For example, if experiment B doesn't produce interesting results, focus on experiment C. Furthermore, as graduate students mature and develop a better understanding of where they and their work are headed, they will increasingly be the ones to set expectations for their work. You will still serve as their advisor, but become more of a colleague.
2. Foster a collaborative work environment
One of the major benefits of working in an academic setting is being surrounded by a community of scholars. While you are your graduate students' official advisor, you should neither want nor need to be their only source of scholarly guidance and discussion. One of the best things you can do for yourself and your students is to foster a collaborative work environment, where everyone can talk about their research with multiple people. This takes pressure off of you, so that projects can move forward when you aren't available to help. It also helps your students become more independent, as they learn to use many sources of information and think critically about many topics. Finally, it leads to better ideas, since multiple perspectives help identify problems and possible solutions.
Regular group meetings are an obvious way to foster a collaborative work environment, where everyone regularly talks about their progress and their problems, and seeks feedback from other group members. If your research group is small, have occasional joint meetings with another group studying related topics. As graduate students mature, put them in charge of training less experienced students. Post-docs should be even better prepared to supervise students.
While graduate students have an intuitive admiration for faculty and post-docs, you should emphasize the importance of treating all members of the department with respect. For example, experienced lab technicians are often more knowledgeable, more skilled, and more efficient in their work than most graduate students. Furthermore, support staff like administrative assistants, technical support, and cleaning staff allow a research program to run smoothly. However, some graduate students feel entitled to disrespect these people. You set the tone for your research group, and should immediately stop such toxic behavior.
Finally, graduate students should understand that they share many resources with other members of their research group, and the larger community. Each graduate student must contribute to ensuring that these resources are ready for the next person. This includes cleaning up after themselves, ensuring that supplies are replaced when they run low, and properly caring for equipment. It may make sense to assign specific tasks (e.g. ordering supplies for X, periodically cleaning out Y) or to have a rotating schedule, but make sure that everybody contributes their fair share.
3. Treat each student as an individual
What works for one student may not work for another. At the beginning of your working relationship with a graduate student—the first year or so—you should constantly assess whether your current method of working together is producing good results. Is the student developing a deeper understanding of the subject matter and acquiring necessary skills at a reasonable pace? Is the student engaged in the project and increasingly taking control of their work? Different people are motivated in different ways, and finding what works well for each student can require substantial trial and error.
In some fields of study, new researchers will interact with their advisor or another supervisor nearly every day as they are learning new skills in the laboratory or field. Some students continue to benefit from a brief daily check-in even after they have learned the necessary technical skills. Other students quickly come to appreciate the flexibility and responsibility associated with less frequent check-ins. Regardless of the field of study, I recommend meeting with new graduate students at least once a week, and balancing the level of supervision that they need with the level that they want.
Also understand that different students have different personalities, different personal responsibilities, and different goals. While some students enjoy listening to music or chatting while they work, others need a quiet environment to focus. While some students will always stay late for drinks and snacks, others leave to spend time with family. While some students are hoping for a job like yours, others want a career in industry, education, communication, or other fields. Respect these differences, and focus on the quality and quantity of work produced, not the number of hours spent in lab.
4. Provide guidance, criticism, and support
As a Ph.D. advisor, your job is to guide your Ph.D. student through the successful defense of their Ph.D. dissertation. You also have an obligation to help your student prepare for the next stage of their career. Therefore, you must help ensure that the student's work meets the high standards of their dissertation committee, of reputable journals, and of possible future employers. You must prepare your students to respond to the harshest critics of their work.
This will often involve pushing your students to do more (e.g. achieving a larger sample size) or to do better (e.g. writing a more comprehensive introduction for their research proposal). It may also involve telling a student that a research idea that they love (and may have spent many months working on) is inadequate for a Ph.D. dissertation. It may be that the preliminary results are unpromising, or that the work would not add anything significantly new to the field. While these may be difficult discussions, it is far better to kill an unpromising project than to let the student continue and end up with work that is inconsequential, unpublishable, and that would leave them poorly prepared for the next stage of their career.
While Ph.D. advisors fully expect to help students with their research, they should also be prepared to help them navigate the "politics" of their department and their field. This may include managing conflicts with people who use the same departmental resources, and the expectations of other professors (e.g. members of their dissertation committee or a professor they are teaching for). Other topics may include requests involving researchers working on a related question (e.g. sharing an unpublished reagent), the expectations of journal publishers and reviewers, and when and how to apply for research funding.
In addition to providing guidance and criticism for your students' research, you need to provide moral support. Good research is challenging, and all researchers encounter failure. Beginning researchers can have an especially difficult time differentiating between a "normal" amount of failure and an "unacceptable" level of failure. By stepping in to provide encouragement (e.g. "This is a common problem/an incredibly interesting question/something I think you'll be able to solve") in addition to actionable suggestions (e.g. "try adjusting B and C as described in previously published work"), you can significantly improve a person's mood and make them more productive.
5. Produce good researchers, not just good research
Almost all Ph.D. advisors are under substantial pressure to publish and to obtain external grants. In the short term, it may seem that the best way to achieve these goals is to treat graduate students as poorly paid employees, who follow your directions to maximize productivity and publications. But if graduate students are not given enough freedom to explore and make mistakes, they will move to the next stage of their careers without sufficient practice in developing their own research questions and recovering from mistakes. This will reflect poorly on you in the long term, so it is in your best interest to produce good researchers, not just good research.