It is a sobering thought, but of the thousands of hopeful students who embark on a PhD this year, many will never finish.
Last year, a report by the Higher Education Funding Council for England (Hefce), revealed that almost a third of full-time and two-thirds of part-time doctoral students had not completed their degree within seven years.
These figures make uncomfortable reading, especially from the point of view of the non-completers themselves. Doing a PhD can be a challenging and agonising experience, requiring enormous personal investment. So to come out with nothing can be devastating.
I know, because I am a non-completer. Over 20 years ago, I started a doctorate in social anthropology at the University of London. But despite working hard for four years, I never managed to submit it and this has haunted me ever since. I have dragged my incomplete thesis with me wherever I go - the piles of notes, books and drafts - in the vain hope that one day I might finally finish. It is written on my CV: PhD (pending submission). Today, my thesis remains in limbo in a filing cabinet upstairs and I am plagued by the feeling of waste - of time, of money, of effort.
In the Hefce report, a breakdown of figures shows that self-funded, mature, part-time students and those in the social sciences and humanities have lower rates of qualification. But there is little understanding of the reasons for non-completion.
I was immature; I chose too broad a topic, my research was poorly designed and I identified too closely with my subject. But, perhaps most important of all, I did not grasp what was required of me and lost my way during writing up.
My experience is not unusual. Dr Cecile Fabre, a senior lecturer and research tutor at LSE, says that underestimating what a doctorate involves is one of two main reasons why students do not complete. Poorly thought-out research design and over-ambition are symptomatic. "The doctorate is a very specific kind of writing and it is not for everyone."
A second major reason for non-completion is lack of funding, says Fabre. "Students who are in financial difficulties have to take on part-time work and simply cannot juggle both."
Dr Christine McCourt, head of the graduate school at Thames Valley University, says that part-time students can find themselves trying to meet impossible demands leading to clashes of commitment that can overwhelm them.
There are things that universities can do to address the problems doctoral students face, and a recent drive to enhance standards in postgraduate provision has pushed institutions in this direction by making them more accountable.
A revised code of practice produced by the Quality Assurance Agency for Higher Education (QAA) in 2004, an emphasis by the funding councils on "timely" submission and qualification rates, and Hefce's plan to publish qualification rates by institution in the future, have all raised the profile of PhD programmes within universities and promoted good practice.
A tightening up of admittance procedures, student monitoring, supervisory training and a move away from the single supervisor system towards a "supervisory team", are all positive moves that institutions have made in recent years.
Sam Griffiths is in a unique position to witness some of these changes. When he embarked on his doctorate in modern history at Oxford University nine years ago, he didn't expect to be dropping out before his three years were up. But not long into the degree, he sensed it wasn't going to work. "I realised that I didn't know how I was going to execute what I said I was going to do." Though he had done a masters, it was in a different subject and he lacked the necessary methodological skills. While suitable courses were available to him, he chose not to take them up. By the end of his first year, he had lost confidence and informed the university he was going to stop.
Four years later, he registered for another PhD, in architecture at University College London, and is now in the final year of writing up. This time, methodological training was a condition of his acceptance on to the PhD programme and it has made an important difference. He was required to shadow a masters course and this gave him practical ways to conduct his research. "That was a big relief - to have a set of 'I can do this' skills."
During his time at UCL, his progress has been strictly monitored. He has to keep a graduate logbook, which tracks student-supervisor relationships, skills strategies and self-assessment. This kind of monitoring can be particularly helpful to part-time students. McCourt says that through providing structure and realistic deadlines, universities can "keep people on track and help balance the demands of the PhD with the demands of the rest of their life".
In practice, the impact of these improvements has been patchy, however. John Wakeford, head of the Missenden Centre for Higher Education, which runs seminars for academic staff and supervisors, receives 100 requests for help each year from PhD students. "Inadequacies in induction, monitoring, records of meetings, feedback and precise criteria for success are often illustrated in their accounts." Wakeford suggests that raising completion rates and attending to the causes of formal complaints might be a good investment for the future.
Nicky Grant, University of London Union vice-president of welfare and student affairs, is not happy with the situation either. She thinks that there should be greater access to funding for students and an improved infrastructure for mature students. Supervisory problems also remain an issue. "Unfortunately, some students are being failed by their supervisors and, ultimately, their institutions."
While most institutions now have training for supervisors, Professor Chris Park, director of the graduate school, Lancaster University, says that the uptake among established staff is limited. "Supervisors' development is critical to the whole process and it's one where the sector is still working pretty hard."
There are some cases where non-completion is not an entirely negative experience. Carli Lestoff started her PhD at LSE in 1988 and worked hard for the first year. In her second year she did some part-time research work in her department, but during her third year gradually sensed she was not going to complete it.
When her funding ran out, Lestoff gained full-time work at the university and now has a good career in social research. But, like me, she cannot quite give up on her thesis. She says she still talks about getting it out of its box and still has the desire to complete it. "I want to finish on the Mastermind principle: 'I've started so I'll finish'."