1. PROCESS AND GUIDELINES

The official tenure process begins in the fifth year, but you should start thinking about it from Day One. There is nothing more daunting than the realization that you should have been aware of each step of the very changeable process you have been participating in instead of considering it only as your tenure approaches steadily in the last year. Some time at the end of your fifth year in the summer--or even the spring--you will submit your CV and you research statement. What you have accomplished goes into your CV. The research statement is typically a two or three page statement of your view of your research endeavors and how your research is relevant to the computer science community. Your tenure case comes under scrutiny in your sixth year. There are several levels to this evaluation. It starts with the department promotions and tenure committee--a committee made up of tenured senior faculty that are elected or partly elected and partly appointed. Then your case makes its way to the department chair. The college level might have a committee; then it goes to the dean. Depending on your university, the case might then go to a university committee, then to the provost or president, and then to the trustees. At each of these levels a letter of evaluation is added to your dossier for the next level's review. And at any of these levels your case might get thrown out. Tenure is evaluated on teaching, research, and service. The importance of these depends on the university at which you are seeking tenure. Prepare yourself for the tenure track by obtaining a copy of your university's promotion and tenure guidelines; typically there will be three of these. There will be one at the university level, one at the college level, and one at the department level. These guidelines will give you a fair idea about your university is seeking in its tenure candidates, and what you need to do to meet those needs. Take advantage of university-wide sources of information. Often, universities will run panels on how to get tenure, and, of course, they will have many accompanying handouts. As well, many colleges and some departments will have their own guidelines that elaborate a bit more on aspects that are important to that particular college or department. Find out what the expectations are for you there. These vary a lot from institution to institution. Find out how much research counts, and how it is evaluated. Find out how much teaching and service count. Don't rely on information about another department in your university; not only can departments vary greatly from campus to campus, but they can vary within the same institution as well. If the guidelines are unclear, talk to your department head or some other person who is qualified to explain the criteria. Your department head and the senior faculty are the people who are going to make the decision at the department level--the most important level. You need a strong recommendation from the department and these are the people who will present that case through the university. Talk to several of these people and find out what they think. The following are standard criteria for tenure evaluation.

1.1 TEACHING

Keep the following matters in mind when you teach. You will become more effective as a teacher--and improve your dossier, if you take teaching seriously and work to improve yourself.

• Consider your philosophy about teaching. What are you trying to do when you teach a course? What is important to you? Some universities require a teaching statement. Even if your university does not, you need to think about your goals and purpose in teaching. Show your commitment to teaching by being aware of issues broader than the scope of a particular course.
• You should teach a variety of courses--but not too many. Teaching a variety of courses will demonstrate confidence in different aspects of teaching. Choose a few courses that you are comfortable with and are interested in and ask the department to give you those courses. You don't want too many courses though; teaching new courses from scratch will inevitably overwhelm you and result in poorer evaluations compared to those you would receive after teaching a course for the second or third time.
  
  Also, don't get pigeon-holed into teaching huge classes all the time. Evaluations from large classes are always worse than those from smaller classes. Appeal to the department chair if you are assigned to beginning classes all the time. Ask to teach upper division or smaller classes. Of course, you don't want to teach these classes exclusively; teach large classes too so they know you're versatile. Your evaluating committees need to know that--at the very least--you can do an adequate job of teaching on all three levels: beginning courses, upper division courses, and graduate level classes.
  
  
• Develop your own course. Developing your own course, whether it is an undergraduate or a graduate course, will indicate to the evaluating committee that you are committed to teaching. Developing special projects--like a programming project or a software engineering project--for your students takes even more time and shows even more commitment.
• Build on the work of others. You don't have to teach every course completely from scratch. Talk to people who have taught the class before. You can get copies of the materials and the syllabus for the class and work from those.
• Get feedback. After the class, ask for criticisms from the students. Don't feel defensive about it. Students will usually give earnest critiques of your performance. When you are accessible to your students, provide help, and are interested in them, your student evaluations will improve.
• Ask good teachers to visit your class. Few college professors have experience with learning how to be an educator, although they spend much of their careers as one. Find people who have reputations as good teachers and ask them to critique your class. Your university may have a faculty development office, or other similar resources to help you improve your teaching. Videotape your class and have someone from this office or a mentor review the tape with you.

1.2 RESEARCH

1.2.1 THE RESEARCH STATEMENT

There is no way you can get around having a good research record. To get an idea of what is needed to construct a good research statement, look at those of successful candidates. Talk with them about those statements. There are also a few standard issues to keep in mind.

• Have a comprehensive plan of research. It is very hard to make an impact on people when you are jumping around between different areas; publishing a few times in several areas means you won't be well known in any. Your publications must be relevant to your overall plan of research. And this plan of research should be reflected in your research statement.
• Produce quality work that is relevant to your statement. Keep in mind that the quantity of published works is far less an issue than is the quality of the work. It is unlikely that an evaluator will have read every one of your papers. Your evaluator will, however, have read a select few of your papers. Make sure your work is coherent; weed out any junk. You have failed to make a good impression if the only paper that is read happens to be your worst, or if it is not clearly faithful to your stated research purpose.
• Be aware that if you pursue interdisciplinary research areas you may be faced with greater difficulty in getting tenure than would someone in a more traditional field. Some research inherently requires collaboration with people outside the department. Unfortunately, this means you have to work harder to get people to know you, because you're covering more ground. You need the support of people from each field involved. You might consider playing it safe by pursuing interdisciplinary research after you have earned tenure at a university.

1.2.2 PUBLICATIONS

Make sure your journal submissions are significant. There are many journals that don't count as much. There are some that count a lot. Put your best work in the best journals in your area. Journal publications are important. Although there are some departments that value conference proceedings very highly, journals are a staple to making your research known.

You also want to be aware of which journals are considered crucial to get published in--i.e., any journal published by the ACM or IEEE is usually considered a top-rated journal. There isn't a standard ranking system for journals, but there are several ways that you can find out which journals are the most important. Look over the authors of a journal; do you recognize them? Are they the people in your field that you respect? More important than the absolute ranking of the journal is how your department perceives its importance. Ask people in your department about which journals are important to them. Also ask your potential letter-writers. Your letter-writers will note in their recommendation that you published in a top journal if you follow their suggestions.

Of course you should not put all of your best papers into one journal. There may be problems with delays and turn-around times; also--depending on the journal--the paper might get published between six months and two years after submission. Remember, though: there are people that want to help you. Do not be afraid of telling the journal panel that you are up for tenure in a year and you would like a quick response to your submission. Often, the editor will send the paper out and get it back in a reasonable amount of time.

Find the right mix of conferences, articles, and grants. Get an idea of how many articles or conferences you are expected to do. Someone like the department chair should be able to tell you this. There is some amount of flexibility in this as well; if you feel that you need eight articles and you only have six, you have not necessarily diminished your chance at getting tenure. It just means you have to have a few more conference papers, or you have to have a strong teaching performance.

Also consider the differences between journals and conferences. Areas of computer science tend to be fast-moving; the problem with journals is that it takes longer to get an article published in them. Conferences, in contrast, publish more timely material. However, journals are considered more scholarly. In many cases a journal will not accept a conference paper that has not been revised.

Electronic journals are important, too. The problem with electronic journals is that at the higher levels of your tenure evaluation your dossier will run into people who are not familiar with the idea of electronic journals. Deans probably will not understand; other faculty might not understand. If you have published in an electronic journal you need to make a strong case about the journal's legitimacy. Show that the review process of the electronic journal is the same as the process for the conventional, respected journal.

Sometimes your research will fall short of excellent. Everyone has papers like that--papers with good, but not great, results (do not submit papers like these to the best conferences). Include the acceptance rates of the conferences you submit to in your dossier. Computer science is a unique field in that these conferences generally have low acceptance rates.

Collaborate. The problem with collaboration is that it is difficult to determine how much you contributed to the paper. Of course, if your collaborators are students, most people will understand that you probably contributed greatly to the paper. If you collaborate with faculty members, you want to make a lot of contributions to the work. Some people start off articles naming themselves as the major contributor. If you decide to make this statement it is crucial that you are justified in saying so, and that you are not slighting your co-authors. You do not want a former collaborator killing your chances for tenure by announcing that you were not the major contributor to a work that you have cited in your CV.

Research independently. Collaboration in research is engaging and fun; however, you want to have an ample amount of works published on your own. You don't want to cut off all contact with your associates, but you want to establish yourself as a researcher who is capable of researching independently.

1.3 SERVICE

Find out how important service is. Traditionally, service is not remarkably important compared to research and teaching. Ask people in your department about service. Find out what is expected.

1.3.1 DIFFERENT TYPES OF SERVICE

There is very little opportunity for pre-tenure level faculty to participate at the college and university level. Most of your service at your institution will be on department committees. Service to governmental agencies is very important. Participation on an NSF panel or workshop counts as service. You should also have a record of contributing to university programs at home and abroad. You will probably be involved with efforts to increase the number of women in computer science and engineering; these are outreach activities that count. Contributing to programs that enhance equal opportunity and cultural diversity also counts. Service to the public is important too. Delivering career booklets to the local high schools and talking to girls about how interesting it is to be a computer scientist or engineer are good ways to participate.

1.3.2 HOW TO SERVE

• Serve on prestigious committees. The best way to do this is to let some friendly senior person in your field know that you would like to serve. Senior people in computer science are often overburdened with program committees and editorships. They simply have too much to do. If they encounter a talented, responsible, and reliable junior person who says,"I'd like to do this," they will probably respond with delight and work to make sure the junior person gets the position. Becoming a journal editor towards the end of the pre-tenure period will indicate to your evaluating committee that you are a capable member of the academic community.
• Choose service with your tenure process in mind. Moderate your volunteerism and choose to serve on committees that help you with the tenure process. For instance, serving on a graduate recruiting committee will avail you access to the students' folders--invaluable information for targeting students to work with you. Or serve as Colloquium chair; as chair you can invite people who you want to write letters for you. You can discuss your research when you are not discussing theirs. These types of service will help you progress towards getting promotion and tenure. There is also service in professional organizations.
• Say no. This might seem simplistic or obvious, but it is a critical point. Women in particular need to say "no" because women are frequently asked despite the dubious benefit of service. You need to establish yourself as a good citizen of the department, but you do not want either your research or your teaching to suffer greatly. Get involved, but not in over your head.

2. HOW TO BUILD A STRONG CASE

Ask for the CVs and the research statements of successful cases. Anybody who has gotten tenure will proud of it. It would actually be more informative if you can get the unsuccessful ones, but these are understandably more difficult to procure. Also keep in mind that expectations change over time. Finding out what the expectations are in your first year and assuming those expectations will hold for another five years is folly. Keep informed about what the department needs or wants. The target may be moving.

2.1 CONSTRUCTING A DOSSIER

One thing to be aware of during the tenure process is that you are evaluated not only by information you provide in your CV during your fifth year, but also by your dossier--which should be a fair representation of your several years at the institution.

2.2 ELEMENTS OF A DOSSIER

2.2.1 RESEARCH

Include your publications in the "research" section. You might categorize your articles by where you've published them, e.g., in refereed journals, books, parts of books, and non-refereed journals. Journals are also divided into multiple sections: "true" journals, conference proceedings where the paper is reviewed in full, and conference proceedings that are refereed by abstract. Those three are evaluated differently by your promotion and tenure committee. Be careful about keeping track of which is which because there are people on the upper level committees that know, for instance, that a particular conference is refereed by abstract; if you fail to separate your articles accordingly they will start to wonder about other papers you have listed. Also include letters of acceptance with manuscripts or papers that have been accepted for publication.

Another area is research funding. Keep track of all proposals and grants that were submitted, funded, are in progress, or are completed. Receiving funding from someone like the NSF, for instance, will be a strong statement for your case, because the deans will know that NSF proposals are peer-reviewed. For the pre-promotion and tenure candidate it is also important to keep track of proposals that were submitted but not funded because it shows an activity profile. It is better to have tried and not gotten funded than to never have tried at all.

2.2.2 TEACHING/GRADUATE STUDENTS

Supervision of graduate dissertations is another element of your dossier.

Try to attract graduate students. Graduate students are a great resource. They can help with your research and your research productivity. Also, when the university evaluates you for tenure, they will inevitably ask: Has she worked with graduate students? How many PhD students has she produced? Where are those students now? Of course, if you do not get PhD students early on in your tenure process, chances are that the students you do have will not be producing work early enough to affect your evaluation. So focus on masters students too. Masters students require more organization on your part because they are going to complete their degrees in two years, as opposed to the five to seven years a PhD student takes. You need to have a clear idea of what you want your masters students to accomplish in that short time-frame.

Of course, advising students can take an enormous amount of time. You are the one that provides both intellectual and emotional support for the student. Sometimes doing research for a PhD can isolate the student; you need to encourage them to continue. Aside from the satisfaction you get from helping these students, the students you advise and work with reflect your efficacy in mentoring and teaching.

Most universities have a teaching award. You need to acknowledge in your dossier that you were nominated for such an award, even if you did not win it.

Keep track of your successes. If you spend a lot of time working on some course development, write it up and put it in your file. Five years later you are not going to remember exactly what you did. Not only should you keep track of successes in terms of what you did, but also keep track of students who are successful. In your sixth year your department head might write to these students and ask,"Can you give me a supporting letter saying this professor was particularly influential in your education?" Letters from students whom you have made a great impact upon carry a lot of weight.

2.2.3 LETTERS

In letters of recommendation, people evaluate your research primarily. They also discuss the impact of your research on the field. They need to be able to say that you have made a significant contribution to a particular area.

Whether your department allows you to choose people to write letters for you or your department seeks out these people on its own, you can cultivate good letters proactively. Attend conferences and workshops. Get as many presentations as you can.

You can also volunteer to be Colloquium chair. As Colloquium chair you can invite all the likely candidates for letter-writing to your university to give a talk. You will have the opportunity to spend the day with them and to show them your vision of research. Then you can invite yourself to speak at their department. You also need to be visible and known to the people in the top departments. Try to give a talk at each of the top ten departments in your discipline within your first three years. Don't be reticent about calling them up and saying,"I'm going to be in town on this date to attend this conference; may I stop by to give a talk?" Few people are comfortable doing this at first, but it is a crucial opportunity for you to create. Soon you will be invited to give talks. Not only will your increased visibility help establish you in the field so that you get strong letters of recommendation, it will help you get funding for your research.

A good time to network is at your interviews; do not decline interviews lightly. Go to universities and make a good impression. When you are asked to interview, the universities and researchers are focusing on you. This is a great opportunity to network and prime potential letter-writers. Ask questions about their work, or about a talk they gave.

If you have produced a paper that you think is relevant to work that somebody else has done, send that person a copy. Don't flood these people with a deluge of your work; choose key works and send those only.

2.2.4 EVALUATIONS

Official written evaluations, which are common to all universities, are a part of your dossier. So the department promotion and tenure committee, the department head, and the dean look at second-year and fourth-year review cases. They prepare letters that are influenced by those reviews. These letters, in turn, become a part of your dossier. When you look at a six-year dossier, you see letters from the sixth year, as well as the second and fourth (or from however often your university has its written evaluations). If one of the committees at one of the levels senses there might be a problem with progress they can make recommendations for special reviews on the alternate years. It is possible for a candidate to be reviewed officially with letters from each year.

If there is no procedure for written evaluations at your university, ask your department head for written feedback. Getting your evaluations in writing is important. It is very easy to sit down with your department head to talk about your progress informally, but your department head might not be as aggressive in explaining the bad--as well as the good--points of your performance. Or you might only hear what you want to hear. You need a written critique to read, understand, and take to heart, uncluttered by the periphery considerations that might sabotage a casual encounter.

When you get your written feedback, expect some candid criticism. No one likes to get criticized, but using this criticism is essential in improving your performance. You might be tempted to put the paper aside for a while because the criticisms are intimidating. This is all right; however, at some point you need to look at it and highlight the points that you can improve on. Make some reasonable goals for yourself and make sure you fulfill them. Future committees will review this written critique and know what points you needed to work on. If you haven't addressed the weaknesses enumerated in the critique, the committees will surely know.

At Pennsylvania State University there is a yearly oral evaluation by the department head. This, in fact, occurs for all faculty--not just pre-tenure faculty. You submit a report detailing all of your activities for the past year. You then have a sit-down meeting with the department head and talk about your strengths and weaknesses, and you outline a strategy for improvement for the next year. These, in addition to the regular written evaluations, will make their way into your dossier.

All universities have some way of evaluating the efficacy of your teaching. This evaluation usually manifests itself as an in-class written survey or questionnaire. Those results are compiled and added to your dossier. The department level committee is responsible for extracting a sampling of comments from these results. And often senior faculty or your department head come in and evaluate your teaching. They write up a form that is included in your dossier.

Do not hesitate to rely on senior faculty for information and feedback. Getting feedback regularly is crucial to staying on track successfully. Senior faculty can review your grant proposals before you submit them, or introduce you to people in the field.

3. THE TENURE DECISION

3.1 EXPECTATIONS AT THE INSTITUTION

3.1.1 MODEL 1: LOW EXPECTATIONS

Be aware of what you can expect from your university. MIT, Harvard, Stanford, and Princeton are all universities where a substantial number of very good people who are hired as associate professors do not receive tenure. This is not necessarily wrong or bad; these universities simply use a model in which a relatively small fraction of tenure-track assistant professors are expected to get tenure. They believe this model builds the kind of institution that they want.

Failing to receive tenure is not in itself a death sentence. These people who are denied tenure are often promoted at other outstanding universities and do exceptionally well for the rest of their careers. However, this model obviously has some negative impact on the climate for junior faculty.

Despite this, there are several excellent reasons to accept a position at such an university. Working at one of these universities might be particularly advantageous if an important person in your field is there.

3.1.2 MODEL 2: HIGH EXPECTATIONS

The second model is that each person hired into a tenure-track position is expected to get tenure as long as they meet the requirements of the university. This is a more common model that universities in the United States and Canada follow. The tenure track process is a little shorter--and thus, harder.

4. CLOSING REMARKS

Although keeping track of all these suggestions might seem overwhelming, it is a reasonable task if you start early and stay organized. Get all the help you need--copies of guidelines, advice from your department head and colleagues, etc. And stay positive. The worst thing you can do is think you're doomed to fail. Work hard, maintain a healthy personal life, and you can do it.