11

I'd like to know what percentage of PhD candidates in Computer Science does not graduate with a PhD and thus drops out of their PhD program.

This number probably varies quite a bit for different countries (e.g., in Continental Europe one usually has to have a Master's degree before being eligible to start with a PhD, while in the US one can "drop out" with a Master's degree) and possibly also for different sub-fields; I'm mostly interested about German speaking countries (Germany, Austria, Switzerland) and PhD candidates with a focus in Algorithms and/or Theory, but any answer "x% in this subfield in that country/at that university" are also welcome.

Background story: Currently thinking hard whether I should go for a PhD in Algorithms at a German speaking university and not a 100% sure if it is the right thing for me. A professor (not the one I'd do the PhD with) urged me to at least give it a try and said that there is always the possibility to drop out if it isn't working out.

Improve this question
3
  • 7
    I don't really see how your question will help you in your background story decision. Knowing whether 30% or 70% of total strangers drop out of similar-but-different programs in similar-but-different countries will yield a very weakly informative prior at best. Better to think about what you want/need to do.
    – Stephan Kolassa
    Mar 30, 2015 at 18:35
  • 1
    It's worth distinguishing between different sorts of dropping out. One Ph.D. student might love their academic work and do very well at it but decide to drop out to run a start-up or accept an exciting and high-paying job in industry. Another Ph.D. student may dislike their academic work, do poorly, and eventually leave to take a pedestrian job they could have started immediately after college. These are radically different experiences, which shouldn't be conflated.
    – Anonymous Mathematician
    Mar 31, 2015 at 2:09
  • While I agree that knowing the information may not be useful for the background story, I disagree that the question is too broad. There maybe a nice comprehensive study out there that has collected and analyized the data. I don't think an answer needs to be a book, nor do I think we need tones of answers (I.e., one for each university and department).
    – StrongBad
    Mar 31, 2015 at 6:23

2 Answers 2

Reset to default
13

The Times Higher Education looked at PhD completion rates in the UK. In the UK, PhD programs are nominally 3 years long, and the Times reported that, averaged across all fields, 70% of students obtain a PhD within 7 years. They predict that 80% of students who start a PhD in the UK will obtain one within 25 years. There are some data for MIT students which shows completion rates asymptote at around 80% after about 8-10 years. There is a fair amount of variability across fields in how quickly the asymptote is reached (e.g., the average PhD in Business is short compared to Medicine), but in all fields about 80% of students obtain a PhD. I was not able to find any numbers for German universities or CS specifically.

These numbers, however, leave out the substantial number of students who stick out grad school because they do not want to "fail". While trying out grad school might sound good in theory, just make sure you are the type of person willing to drop out if it is not right for you.

Improve this answer
2
  • "leaver out the substantial number of students who stick out grad school" - so they just stick around until eventually a PhD certificate finds its way on their desk?
    – mort
    Mar 30, 2015 at 14:01
  • 14
    @mort the people who are "sticking it out" continue to work at their PhD, often harder than the people who want to continue in research, but they tend to hate (maybe too strong of a word) a good portion of their life. It is kind of like training for a marathon. Some people give up on their first training run, others enjoy the training and the race, despite it being hard work, and go on to do other marathons, and a third group complete the training and run the race despite hating it. Most of the third group never run again, but a few change their mind when they get to the finish line.
    – StrongBad
    Mar 30, 2015 at 14:11
1

The completion percent will vary between universities.
For USA, I can introduce you to the NCES ( National Center for Education Statistics). All univeristies are required to report information each year about the university students, faculty, salaries, demographics, degrees, retention, etc.

There are many data tools within NCES. You might find the IPEDS tool (Integrated Postsecondary Education Data System) most helpful for your question. It allows you to pick specific universities and examine data. I do not know if the countries you mention have similar data collection systems.

Many students complete the course work but do not finish the degree. This is know as All But Dissertation (ABD). I don't know if you want to count these folks as they haven't officially dropped out. However they have not ( and several will not) complete their dissertation.

Improve this answer
1
  • As far as I can tell, the needed data are absent. At the University level I can see how many graduate students enrolled in a given year and how many graduate degrees were issued in a year, but some students may get multiple graduate degrees (e.g., an MD-PhD student who gets an MS and MPhil along the way). Ignoring that, I would still need to track incoming students and outgoing students over a couple of year period. While there is some field specific graduation data, there is no field specific data about number of incoming students.
    – StrongBad
    Mar 30, 2015 at 14:25

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .