John Holloway,
Professor of Chemistry and Geology

Mentoring for Student Survival in an Experimental World

Mentoring is a complex, interactive process between advisors and their students. The process is strongly influenced by the nature of the research and by the stages that the students pass through during their career in graduate school. In the following narrative, I begin with a brief description of our research and then describe some critical points in the graduate school experience and my role at those times. At the end I will include some of my guiding principles.

The goal of our research is to understand the conditions found in the interior of the Earth and other Earth-like planets. Our purpose is to understand the nature of the minerals below the planet's surface. For instance: What are their crystal structures and chemical compositions? How do the minerals react with each other with changing pressure (depth) and temperature? What are the conditions that cause rocks in the Earth's interior to melt? And, how do those melts form magmas that may reach the surface as lava flows or explosive volcanic eruptions? This is "pure" research, much of which is a fairly long way from finding use in practical applications. However, some of our work is directly applicable to understanding volcanic eruptions, and some of the techniques we use are finding immediate application in materials science. My former students have found jobs as faculty members at research universities in the United States and abroad, and as research scientists at national laboratories, at a Carnegie Institute laboratory, and at corporate research labs.

The lengthy road to obtaining a doctoral degree begins with choosing a university, department, and advisor. I first try to influence students' decisions by actively recruiting them. I recruit new graduate students by giving talks at other universities, at national and international meetings, and by networking with colleagues around the country and the world during sabbaticals and other leaves when I work at other research universities. For instance, of my first three Ph.D. students, I met one in Tucson, one in Edinburgh, Scotland, and one was recommended by a close colleague at the University of Cincinnati. My goal in recruitment is to get good students to come to our graduate programs at ASU. However, I do not recruit them specifically to work with me because it is terribly important for students to carefully select a research advisor.

Such caution is needed because the relationship between advisor and Ph.D. student is one of the closest, most intense, and stormiest of all interpersonal relations one has in a lifetime. It has often been described as a combination of a parent-child and marital relationship. The reason behind this is the high expectations that each individual has of the other, expectations that are invariably too high on each side in the beginning.

After a student and I have chosen each other, the mentoring process begins with advice on courses to take. Graduate school is the last exposure students will have to formal course work. Because geologists in the future will rely on advanced concepts in chemistry and fluid dynamics, I ask my students to take courses in thermodynamics, quantum chemistry, and fluid dynamics. These are challenging courses, but research being done by my colleagues at ASU illustrates how these fields can be applied to problems in geology, so it is relatively easy to convince students to take the courses. Once enrolled it is important to provide students with support and encouragement, for at this point students are usually feeling insecure because they are surrounded by more experienced students who appear much more knowledgeable.

Adding to the insecurity is the comprehensive exam, which is taken in the third semester. For the exam the student must write two research proposals, one of which usually becomes their dissertation project and one which is an original idea not related to work they or their advisor is doing. Two weeks after presenting these proposals to the five-member examining committee the student defends them in an oral exam. Before and during the development of the proposals it is my job to encourage the student's creativity as much as possible.

Encouraging creativity is the most important part of mentoring; all the rest is just making sure that a student will have the necessary tools to implement their creative ideas. Most of the academic process from grade school on seems designed to destroy creativity, so the creative spark is pretty fragile by the time a person reaches graduate school. There is a paradox here because a lot of creativity begins with questioning authority, in this case scientific authority. But the advisor is viewed by the student as a prime source of scientific authority, so part of the job is to get students to question me and my view of the state of my discipline. I often find myself trying to encourage ideas that I do not personally believe to be correct, but I try to remember that all new ideas in science were considered incorrect by the practicing scientists of the time.

Finally, the student takes the oral exam. This is the first and perhaps the most important milestone in a graduate student's career, and an enormous amount of anxiety is attached to the exam. Students are very much on their own in preparing for this exam, unlike future hurdles, such as planning a research program, presenting a first paper at a national meeting, or writing the dissertation. During the exam the student tries to impress all of the committee members—especially me, the advisor— and I am anxious, hoping that the student will do well both for their sake and for my own sake. One of my foreign students failed his exam the first time, primarily because in his native culture, questions are only answered if the answer is unambiguously certain. Many of the questions he was asked could not have certain answers and he declined to answer them. This cultural clash took me by surprise, and I spent the next several weeks explaining our approach of answering questions in the best way possible given the available data. He passed on his second try, finished in four years, and is now on the faculty of a university in his own country where he has built his own lab and is publishing important papers in international journals.

Once the comprehensive exam is over, the student and I get down to the serious job of refining the research topic and research plan. It is an old adage that it takes just as much effort to work on a trivial question as an important one, so I try to focus student's interest on important problems. As I mentioned earlier, we attempt to simulate the pressure-temperature conditions in planetary interiors in order to understand the chemical and physical processes that result in partly melting particular regions in the planet. We generate the high pressure and temperature conditions with several kinds of mechanical-electrical devices. To do research using these machines, graduate students must learn to be plumbers and electricians. The samples of rocks, minerals, and volcanic gases that go into the machines are very small and must be placed in delicate gold or platinum capsules which must be sealed by welding, so the students must also become jewelers. These skills are not taught in courses but learned in an apprenticeship system in the lab.

This program takes the form of a research group consisting of myself, a postdoctoral research fellow, and two to four graduate students. New graduate students learn from the more experienced students, the post-doctoral fellow, and myself. In the research group there is almost daily contact between all members of the group. This close contact makes it easy for me to listen to students' problems, ideas, and goals, and to make short-term suggestions to help them survive the ordeal of graduate school. In addition to the informal daily contact in the lab, the research group holds weekly meetings at which each person's current research is discussed. These meetings also provide a venue for practice presentations of talks that will be given at professional meetings.

Often I have been fortunate to have visiting scientists in my research group. These are usually colleagues on sabbatical leave from other universities. During one memorable period visitors from Australia, France, and Scotland were here for six months. Other scientists visit for shorter periods of time to learn specific research techniques, often from the graduate students. Each of the visitors provides the students in the research group with a different perspective. The foreign visitors also demonstrate that excellent science is being done in other countries. Another positive benefit is that the visitors get excited by the students' research and provide strong encouragement for the students to keep working in the face of what often seem overwhelming odds when experiments continually fail.

By the second or third year of their graduate career students will have chosen a research project and begun experiments. To answer an important new question usually requires developing a new experimental technique, modifying an existing apparatus, or both. The students have probably never done either of these before, but I push them to do as much as possible on their own, working out equipment designs, searching through suppliers' catalogues, learning the tortuous task of placing a purchase order, and constructing equipment from the parts. Students have an initial advantage if they have rebuilt old cars or taken a machine shop class in high school; these are typically male experiences, but not all of the male students and very few of the women have had this training. To overcome this lack of experience my students take an informal machine shop course run by the mechanical shop, and I do a lot of informal instruction as we work in the lab. This training works well; one of my students, Dr. Connie Bertka, is now a staff member at the Carnegie Institute's Geophysical Laboratory, one of the world's most prestigious research labs in the earth sciences.

Regardless of background, getting an experimental apparatus to work properly is difficult, and after many failures a student questions why they have chosen such a frustrating endeavor. My job at this point is to try to keep them from becoming, not always successfully, psychotic alcoholics. This is another instance in which the research group is a big help because younger students will have watched the rest of us having similar problems, and some successes. The example of successful experiments is important because in the worst cases more than a year of failures can occur. Throughout these difficult times I try to demonstrate how much fun it is to be wrestling with these challenges, especially when you finally win.

Eventually the experiments do work satisfactorily and results start coming in. Often the results are difficult to interpret, and here is where many of the physical and chemical principles learned earlier are put to good use. Sometimes we agonize over interpretation of results for weeks or months, doing more experiments to test our interpretations. This is another tricky time because, having learned to make the experiments work, students show a stronger preference for doing more experiments rather than trying to understand the ones they have done. At this point I sometimes exile them from the lab until they have an answer. When the answer comes it is time to begin writing up the results in the form of a paper. In today's job market it is a big a dvantage for a student to have at least one research paper (and preferably two or three) finished by the time they are looking for a job, so I try to implant the thought of publishing at an early point in a student's career.

Another important skill is writing proposals to obtain research funding. My approach is to have students help by writing a summary of the research they have done. These summaries form an important part of proposals and students can see the connection between their work and the continuing success of the lab. In the course of their career in the lab, students usually have helped write three or more proposals.

As graduate students mature professionally, they need to make a transition from advisee to independent researcher. This transition is not an easy one and is spread out over the duration of time in grad school.

I believe that there are four stages in the advisor-advisee relationship. The first is the euphoric honeymoon in which the student and I think we have found the best advisor and student possible. The student thinks I do the most perfect science around, know everything about my field, and will smoothly convey all of this to them. I think that the student has a terrific background, can effortlessly master the most difficult classes, and learn all of the lab techniques in the first year. The second stage is cascading disillusionment beginning with the student detecting subtle errors in my judgment of course assignments and ending with the certainty that the experiments we mutually chose in the beginning are practically and theoretically impossible to accomplish, and that they made a disastrous mistake in choosing me as an advisor. In the meantime, I perceive that the student's background was drastically overrepresented, resulting in such difficulties with their course work that they won't learn anything. It seems that the student cannot focus on a research topic, is all thumbs in the lab, and that I have made a disastrous mistake in taking on this student. The third stage is grudging acceptance. This occurs sometime after the experiments have begun to work. The student sees that on a few occasions I have a helpful idea and that some of the course work I insisted on is useful. I notice that the student has independently come up with one or two good ideas, has done very well on the comprehensive exam, and has actually learned to operate some of the machines in the lab. The final stage, mutual appreciation, is reached fairly late in the student's time in the lab when we are actively working on papers together and realize that we share common goals and knowledge.

After the first few students I realized that this was a general progression and tried to limit my overreaction to each new student. For students, however, this is an entirely new experience and I can only partially convince them that all will be okay in the end.

So how do these descriptions of the process of the Ph.D. translate into a mentoring philosophy? Let me state some of my basic tenets. First I treat my graduate students like the young professionals they are. It is important to remember that they were in the top of their graduating class and have a bachelor's degree. If they were working in the corporate world they would be at a junior manager level with their own secretary, telephone, and a lot of responsibility. Few people at a university have their own secretary, but I give a student a lot of responsibility, coupled with the authority to make decisions on major and minor purchases, and on equipment scheduling. I expect them to work out problems on their own, providing just enough advice to keep them going. I spend a great deal of time telling them that making mistakes is inevitable, common, and necessary in the course of doing science. The more innovative and original a research project is, the more mistakes will be attached to it. As I encourage students in their experiments I try to eliminate the fear of failure. I am not perfect at this and may tell students either to do a certain experiment that they think shouldn't be done, or else not to do what they want to do. The mark of good students is that they ignore these suggestions of mine, and go ahead and do what they think is right; but they eventually have to defend what they did. I encourage them to think broadly about how the research they are doing fits into science and the human endeavor, but also to focus very narrowly on their own project. I expect my students to be the world expert in their small corner of science by the time they finish. I then expect them to quickly broaden their view when they move on. Throughout all of this experience I try to demonstrate that doing this kind of research is more fun than drudgery and that we should feel extremely lucky to be able to do it at all.

The combination of this approach and the generally excellent students I have had seems to work; of the fourteen Ph.D. students that have started with me, twelve have finished in an average time of between five and six years. Two have not finished. Of the twelve, all but one are employed doing the kind of research they learned with me at Arizona State University.