What to expect in a ph.d program


In this article, I will talk about the seven stages of a Ph.D, and the typical duration of a Ph.D. Let me start this discussion by saying that no two Ph.D programs are alike. The nature of the Ph.D program depends on the career aspirations of the student. Students who aspire to join academia might want to go more into exploratory research, whereas students with industry in mind might want to focus more on practical aspects. Here are a set of possible options:
  1. Fresh out of B.Tech directly joining the Ph.D program
  2. Student is joining Ph.D after a Master's (M.Tech or M.S)
  3. Part time student
  4. Student is on leave from an academic institution
  5. International student
It would be prudent to believe that students in these five categories will have radically different career aspirations. Consequently, their expectations from the Ph.D program will be different.
Let us now break a typical Ph.D program into several phases, and see where it needs to diverge for different students having different expectations.
Here are the different stages of a typical Ph.D program.
  1. Relevant course work
  2. Skill development
  3. Study of relevant technical literature
  4. Participation in a project led by a senior student
  5. Research effort : prototyping ideas + writing papers
  6. Mentoring junior students
  7. Helping the professor in formalizing future research directions
These seven stages are supposed to take a student from a novice level to the highest level of academic attainment, i.e., guiding others to perform world class research. We can divide them into broadly three stages namely Education (1 - 4), Research (5), Leadership (6-7). The most interesting part is that research is just one stage out of seven in a typical Ph.D program. This might come as  a surprise to many; however, this is fortunately or unfortunately true. On a better note, this division into stages, does not mean that time is equally divided across the stages. In fact, it is possible to pursue a set of above mentioned goals simultaneously. The research component might be 99% of the entire Ph.D effort, or might be even be as low as 30%. This is dependent on how long the previous stages take. Secondly, all the stages are not compulsory for everybody. It is theoretically possible to directly jump to writing research papers (Stage 5) and leave out everything else subject to institute rules. Let us now look into the stages in some more detail.

[1-3] Courses, Skills, Literature Survey
Please refer to the figure in the link, which shows that a Ph.D is a very humble and modest effort in extending the state of the art. The objective is to create something, which is completely new, technically sound, and desirably world class. To propose something that has never been proposed before, a student needs to be able to understand extremely complicated research papers, figure out problems in them, and look at ways to solve them. Achieving all of this, requires a very good knowledge of fundamentals as well as superlative language skills. If a student is extremely good technically but lacks the knowledge of English that is required to understand a research paper, then he will not be able to conduct research fruitfully. A building is only as strong as its foundation. The foundation is never visible, but it holds the building together, and is often blamed for catastrophic failures.
In the US, kids typically write fairly complex software in school like small computer games. They move on to fairly complicated software like iPod or Android apps in college. These are also very decent means of making money. Secondly, since English is their mother tongue, they don't face any language issues. Unfortunately, we don't have these advantages in India. Hence, it becomes essential to invest on programming skills and linguistic skills before we undertake any project.
We should understand that one very important goal of Ph.D is getting a much better job. This basic education takes us a long way towards that goal. After the end of this goal, we would expect a student to be a fairly proficient programmer. He should be able to undertake fairly complicated tasks like modifying the linux kernel, or the gcc compiler without a lot of difficulty. Secondly, he also should have developed a good knowledge of the relevant literature in the area. Lastly, his language skills should have developed to a point where he can write a piece of text that can be appreciated by an international audience.
In American universities, there is an exam called the "Ph.D qualifier" that checks for all of these skills. Students are denied admission to the Ph.D program if they fail to pass this test. In India the qualifer is not that strict. My philosophy is that students can and should be coached to reach the desired level. It is very essential to ensure that a student meets all the quality standards before he can embark on research; otherwise, research will be a very frustrating endeavor.

[4] Working with a Senior Student
A typical question that comes to a student's mind is : if my Professor gives me something to solve, how do I solve it? How is research done? This is a skill that takes some time to develop. There is no reason to rush on this, or show haste. With time and perseverance, anybody can become a great researcher.
One quick way to learn how to do research is to work with a senior student, who is already doing a lot of research. Younger students can learn a lot of hands on tricks, short cuts, and useful trivia from a senior student. Senior students typically make the best mentors and guides for newcomers.

[5] Performing Research
Now, here comes the real meat of the Ph.D. Once a student is well trained, and has the relevant experience, it is time for him to start working on his Ph.D research. To reach this phase, it might take anywhere from 3 months to 3 years. It depends upon the student's background, and his desire to learn. It is my firm belief that if a student is desirous to learn and work hard, he should be able to overcome all kinds of obstacles, and quickly become world class.
The most popular myth about a Ph.D is the phrase "Ph.D problem". In a Ph.D, the main problem is the supervisor wink, who keeps on urging students to work. The rest is easy.
There is more about my model of research in this site. To summarize, I will start out by trying to align my interests with the student's interest. Once a match is found, we will quickly finish a survey of some related work, and review some of the areas and problems that I have been working on. Most of these problems will typically also have a corporate sponsor. The student is expected to keep them aware of his research. This will ensure that his research is relevant, and he can get good internships(preferably overseas), and job opportunities in the future.
Once the student choses an area to work on, we will get together and have intensive meetings to find out shortcomings of previous work, and where we can contribute. This will be the most intensive part of the Ph.D. After having discussions, the student will begin the task of implementing the research idea. This will take around 3 to 9 months. Once, we get some good results, we will summarize them in a research paper and send it to an international conference or a journal. It will take some time for the paper to get accepted. It is typically necessary to iteratively refine the work. Finally, the research paper should get into a reputable conference/journal. I would ideally prefer that the student travel overseas to present the paper in a conference as frequently as possible. Please note that this is contingent on the availability of funds. A Ph.D will typically have two or three papers worth of work.
During the course of the Ph.D, it is prudent to expect the student to become more and more independent. Towards the end of this stage, he should be able to find out problems of his own, and work on solutions.
[6-7] Mentoring Stage - new research directions
 Once a student has finished about three fourths of his Ph.D work, it is time for him to achieve a bigger role. He needs to be given the opportunity to lead others. This is very important for two reasons. Firstly, if he wishes to join academia, then he is getting an opportunity to mentor students. If he wants to join industry, then he can show this experience to get directly hired as a team lead or manager. It also greatly broadens his scope and challenges him to do even more innovative work.
I would also expect students to take a look at real world research proposals that are submitted to government bodies and other funding agencies. He can also participate in them, and also work closely with corporate sponsors to learn more about their work. This will prepare him to take on important roles after graduation.

Different categories of students
Here are my views about the 7 stages of a Ph.D, and which stages are optional, and which stages are compulsory.

courses
skills
lit.review
work with
senior student

research
mento-
ring
setting
research
direction

B.Tech
yes
yes yes optional
yes

M.Tech

yes yes
yes

Part
time
yes yes yes
yes

College
teacher

yes yes
yes optional optional
Intl.
optional yes yes optional yes

yes means that it is highly desirable
optional means that it is advisable
The rest of the entries are truly optional. Please note that students will have to complete a minimum set of requirement imposed by the department.
Here is another table based on the career choice after Ph.D.

courses
skills
lit.review
work with
senior
student

research
mento-
ring
setting
research
direction

Academia
yes yes yes yes yes yes yes
Research
Lab
yes yes yes optional yes
yes
Industrial
R & D
optional yes yes
yes yes
Start-up
yes yes yes
yes yes
Others
optional yes yes
yes


How long will a Ph.D take?
This is needless to say the most important question in the mind of a student. A lot of people scoff at the idea of a Ph.D because of the time commitment. I want to assure you that if you are the Ph.D type, then this time is well spent. Nonetheless, let us try to answer this question.
An accurate answer to this question is - whenever you will be able to convince the Ph.D committee that you are ready to graduate. The Ph.D committee typically consists of other faculty members in the department, an external Indian reviewer, and an external foreign reviewer. Whenever, we are sure that all of them will approve the thesis, the student can graduate if he wishes to.
The next question is : When will these people approve a thesis? This needs to be answered by taking a look at all the students who have graduated in the past. Once a student's contribution is roughly the same as the previous students, he should be confident to get through.
Ph.D programs typically take about four to six years, depending upon the performance of the student, and his desire to graduate. If the student is a low performer, then he will take longer. Sometimes high performing students desire to graduate sooner because they might have family commitments, and need to settle down soon. Surprisingly, many a time it does happen that students decide to stay a year or two longer to strengthen their resume, such that they can get a better job. I am fine with all of these configurations, as long as the student's thesis is worthy of getting approval from the Ph.D committee.