PDA

View Full Version : "Accelerated" PhD - How do people do that??



econornot
11-21-2010, 11:20 PM
I think it's fair enough to say that the typical time needed to complete a PhD these days is 5 to 6 years, typically consisting of 2 years of coursework followed by 3 to 4 years of research on one's dissertation. But every year, we do see some people being in the job market when they are only in their 4th year (and it really isn't rare) and every now and then, we see people completing even in just 3 years. (And I am not referring to people like Glen Weyl!)

I am just curious to know, where is the part of the program that is possible to be or usually "accelerated"? Is it the first year coursework part where some people who, perhaps, had already done a master or are just too smart that they are able to take and pass qualifying exams immediately upon entrance and skip their first year of coursework... or do some manage to squeeze 2 years of coursework into 1.... or is it always the case that they simply just don't need more than a year to write up their dissertation.
I have only personally known 1 person who completed in 4 years and it's just that she spent only 2 years on research, but I do not have any personal experience on any other cases. Everyone else I know took at least 5 years.

Anybody care to share their experience or insights??

SlowLearner38
11-21-2010, 11:43 PM
See, I look at it from the opposite perspective. I don't understand why it takes so many people 5-6 years to finish. I think a surprising amount of people start their PhDs with too little understanding of how the game works. They don't make decisions about research until someone tells them to. To me the logic is simple: The PhD is a research degree, so I should be trying to start research as soon as possible. People don't start thinking about research until too late and they don't think ahead of time that if you want to finish in 4 years, you need to be done with your research by the START of your 4th year. That just keeps pointing to the fact that the people who finish in 4 years are the ones who realized these things early on, so they begin research as soon as possible.

charlesr
11-21-2010, 11:58 PM
My guess is that the dissertation research time is the significant variable. I'd be surprised if many people skip the first year coursework.

kevinp123
11-22-2010, 12:28 AM
Definitely agree with whats been said already. I think most people who finish quickly already have an idea of what they're interested in before they start and may or may not have already begun research. One thing I do now that will hopefully pay off is write down any dissertation idea I have. If you ever notice something and you're not sure how it work or think it would be an interesting dissertation topic, write it down so that you can look into it later. Then when it comes time to sit down with you're advisor, hopefully at least a couple of the ideas you have written down will be suitable for a dissertation topic, or at least a start.

asquare
11-22-2010, 12:29 AM
See, I look at it from the opposite perspective. I don't understand why it takes so many people 5-6 years to finish. I think a surprising amount of people start their PhDs with too little understanding of how the game works. They don't make decisions about research until someone tells them to. To me the logic is simple: The PhD is a research degree, so I should be trying to start research as soon as possible. People don't start thinking about research until too late and they don't think ahead of time that if you want to finish in 4 years, you need to be done with your research by the START of your 4th year. That just keeps pointing to the fact that the people who finish in 4 years are the ones who realized these things early on, so they begin research as soon as possible.
With all due respect, let's see what you say in a few years ;)
As you note, in order to graduate in four years, you really need to finish your job market paper in just over three. Your job market paper needs to be written and pretty polished by October of the year you are on the job market, and it needs to be ready to send out to prospective employers by November of that year. So, going on the job market your fourth year means essentially writing your job market paper during your third year, or over the summer after your third year. You also need at least one other paper more or less complete, because employers will be reluctant to hire someone who they don't believe will actually have their degree in hand by the time the position is due to start. If you still have two chapters to write when you are on the market, it's hard to convince employers you will really finish that spring/summer.

You can and should start doing research early. Work that you do for your 2nd year field classes should contribute to your dissertation in some way, even if it is basically a literature review or even a proposal for a project you begin later. But actually having a mature research topic by 2nd year or earlier is unusual, even for students who come in motivated to do research and finish quickly. The material you cover first and second year, and the seminars you attend, do contribute actually to the research process. It's a rare student who is able to begin high quality work that leads to a job market paper immediately, without the benefit of some time in graduate school. The research most students are prepared to do when they begin graduate school frankly isn't good enough for a job market paper -- and that's the way it should be! The PhD program is meant to teach you to do research, and that takes time. Rushing out of the program in four years isn't a good strategy long term if it results in a lower quality placement or simply less learning while in graduate school. Taking advantage of the mentoring available to graduate students and the seminars and other resources at your school, as well as the time to focus on your research, is important and shouldn't be undersold.

econornot
11-22-2010, 12:46 AM
I tend to agree with asquare on this. I am in my 3rd year and how me and my supervisor worked to finalize a topic was by meeting regularly and discussing a few papers in the general area of our interest each time, post a few questions and comment on what we like and not like, which eventually led to a topic. I personally find it terribly hard to spend so much time on reading so many papers during my coursework period... you can put it down as a lack of motivation or just not brilliant enough to handle it, but I believe if (most of) the students have spare time to do these things during the coursework period, then the program could have been made tighter.
But then, i still beg the question... how do people do it???? The few post above tend to suggest it's a case of shorter research period, but honestly, I tend to believe that it's the case of people skipping coursework and saving time there. The reason is that people who finished PhD faster almost always only come from the US, and to my knowledge, it's possible to simply go straight in and take the prelims in the US if you are good enough, whereas in Europe, the coursework phase tend to lead to a full master degree and no matter how brilliant one is, he or she still has to go through the motion of attending classes, etc. But then, it could well be the case that the most brilliant people are in the US schools and only these people are capable of finishing their PhD faster than others.

Any views on that?? I am pretty sure some of you have got first hand experience of people in their faculty being one of those 3 or 4 years genius!

Elliephant
11-22-2010, 01:08 AM
I know a guy who finished his PhD at Berkeley in four years. He did the standard two years of coursework, but because he had RAed for two years before starting grad school, he came in with a well-defined research question and was able to write his dissertation in two years. However, I'm not sure that worked in his favour. He ended up spending two years as a post-doc before landing a faculty position. I would guess that this was partly for the reasons given in asquare's second paragraph.

taurenchieftain
11-22-2010, 04:08 AM
I think if you know what you want to do, and is going to do empirical work. You can in thoery start your work NOW. Lots of good empirical work can be done without knowledge of the PhD courses. Another thing is that, if you are doing empirical work, you can start preparing your data set now. In some fields, it may take you a year to get, combine, and correct the errors (through manual searching etc) in your dataset.

I have heard of people graduating in three years without some special circumstances. The usual case is that you did your undergrad in the same institution, where you basically took all the first year phd level courses or more, and you did research with your RA and know the faculty there well. But again, those people basically have their dissertation committee half set up the minute they enrolled.

Essayvision
11-22-2010, 04:10 AM
Most phd programs end in minimum 5 years.. 4years is very tight although i know people who's done it. That requires finishing courses in one year (depending on type of program) and if you start on your thesis right away and be able to get your paper published by your 3rd or 4th year, you can definitely aim to finish within 4 years.

asquare
11-22-2010, 04:15 AM
I think if you know what you want to do, and is going to do empirical work. You can in thoery start your work NOW. Lots of good empirical work can be done without knowledge of the PhD courses. Another thing is that, if you are doing empirical work, you can start preparing your data set now. In some fields, it may take you a year to get, combine, and correct the errors (through manual searching etc) in your dataset.
Sure, the mechanics of analyzing data don't require PhD coursework in many cases. But identifying an appropriate research question, identification, and data set, and choosing the right methods to analyze the data, absolutely benefit from the coursework and other interactions that take place during the first two-three years of the PhD. Research that is started before graduate school or during the first year is unlikely to be mature enough to become part of the dissertation or the job market paper. Getting started early is great and will provide valuable experience, and I don't discourage it. However, there really is a big difference between the empirical research that an advanced undergraduate or first year PhD student produces, and the empirical research that a 3rd or 4th year PhD student produces.

asquare
11-22-2010, 03:35 PM
Most phd programs end in minimum 5 years.. 4years is very tight although i know people who's done it. That requires finishing courses in one year (depending on type of program) and if you start on your thesis right away and be able to get your paper published by your 3rd or 4th year, you can definitely aim to finish within 4 years.
I'm sorry, but this simply isn't accurate. Very few economics PhD students have publications before they finish graduate school, and publishing a paper is definitely not required in order to graduate. Also, it would be extremely unusual to finish courses in one year, even for students who do finish their PhD programs in 4 years.

Essayvision, I have to ask -- do you know anything about graduate school in economics specifically? Are you especially interested in economics? Why the flurry of posts in this forum?

SlowLearner38
11-22-2010, 04:39 PM
With all due respect, let's see what you say in a few years ;)
As you note, in order to graduate in four years, you really need to finish your job market paper in just over three. Your job market paper needs to be written and pretty polished by October of the year you are on the job market, and it needs to be ready to send out to prospective employers by November of that year. So, going on the job market your fourth year means essentially writing your job market paper during your third year, or over the summer after your third year. You also need at least one other paper more or less complete, because employers will be reluctant to hire someone who they don't believe will actually have their degree in hand by the time the position is due to start. If you still have two chapters to write when you are on the market, it's hard to convince employers you will really finish that spring/summer.

You can and should start doing research early. Work that you do for your 2nd year field classes should contribute to your dissertation in some way, even if it is basically a literature review or even a proposal for a project you begin later. But actually having a mature research topic by 2nd year or earlier is unusual, even for students who come in motivated to do research and finish quickly. The material you cover first and second year, and the seminars you attend, do contribute actually to the research process. It's a rare student who is able to begin high quality work that leads to a job market paper immediately, without the benefit of some time in graduate school. The research most students are prepared to do when they begin graduate school frankly isn't good enough for a job market paper -- and that's the way it should be! The PhD program is meant to teach you to do research, and that takes time. Rushing out of the program in four years isn't a good strategy long term if it results in a lower quality placement or simply less learning while in graduate school. Taking advantage of the mentoring available to graduate students and the seminars and other resources at your school, as well as the time to focus on your research, is important and shouldn't be undersold.

It would be hard for me to say that you don't make great points here. I guess at the end of the day, graduate students are extremely heterogeneous with respect to their abilities, goals, motivation, and so on. I agree that the research a 1st year student produces probably isn't going to be good enough for a job market paper. However, that doesn't mean the research is a complete wash and waste of time, because it most certainly isn't. The job market paper may be on the exact same topic as the one you wrote in your 1st year, but by the time you get to the job market, you'll have improved the quality of the paper and the insight taken. Point being: 1st year students are capable of having good research ideas, but obviously they won't have the complete set of tools needed to appropriately answer the question. I guess the big thing is that starting research early gets the research wheel turning, so by the time you're on the job market, that wheel is greased and moving with minimal friction.

econornot
11-22-2010, 04:46 PM
Very few economics PhD students have publications before they finish graduate school, and publishing a paper is definitely not required in order to graduate. Also, it would be extremely unusual to finish courses in one year, even for students who do finish their PhD programs in 4 years.


The having to publish papers before graduation is obviously crap since the refereeing in econ journals can take years so it's only possible if one submits a paper by say second year. But I am, however, under the impression that our phd dissertation must be at very least of publishable standard... am I not right on that?

Also, I am very curious on how many people actually finish the courseworks in 1 year, or perhaps, in some form of accelerated manner. I see a number of schools offering incoming students the chance to sit for the re-take exam meant for the previous year students, and if a person had done a master from a decent place or is a transferred student (in which both cases are not exactly that rare), it is highly possible to pass the prelims (especially in micro where almost everywhere's micro sequence tend to be MWG + contract theory at Bolton Dewatripont + game theory at Fudenberg&Tirole standard). If somebody does managed to pass one or two sequences, what does he usually do?? Take the field courses early, or simply enjoy a lighter courseload (and perhaps have a bit more time to start thinking about his research).
I have a friend who's doing a phd in the US, and he actually passed 2 exams (micro and metrics) and was allowed to be exempted from them. But it appears that the school does not have a proper structure of such "acceleration", and he doesn't know what to do after it, and he ended up attending the micro and metric sequences again, for firstly, he doesn't know how he could better spend his time and nobody in his dept seems to want to commit to any advice to him, secondly he is afraid that he might not have learnt sufficiently even though he has managed to pass the re-take prelim (and he is an applied micro guy), and thirdly, he was the only student in his cohort to be exempted from anything and he simply didn't want to be left out from the rest of the cohort. I don't know what you guys think, but I find that really silly.

econornot
11-22-2010, 04:51 PM
Point being: 1st year students are capable of having good research ideas, but obviously they won't have the complete set of tools needs to appropriately answer the question. I guess the big thing is that starting research early gets the research wheel turning, so by the time you're on the job market, that wheel is greased and moving with minimal friction.

I would say that 1st year students being capable of having good research ideas is as rare (or common if you like) as them having the complete set of tools to answer the question at a appropriate level and rigor. I believe it takes time to learn what is a good (or even just relevant) research question or idea; while creativity is important, it is by no means near a sufficient condition to be able to come up with a decent economic question.

asquare
11-22-2010, 04:53 PM
The having to publish papers before graduation is obviously crap since the refereeing in econ journals can take years so it's only possible if one submits a paper by say second year. But I am, however, under the impression that our phd dissertation must be at very least of publishable standard... am I not right on that?
Yes, the dissertation should be of publishable quality. Standards vary depending on degree granting university and some students will have higher quality dissertations than others, of course. Not every dissertation chapter winds up published. But that's the goal. Usually, though, dissertation chapters are submitted after graduation. It's certainly not necessary that they be submitted before.


Also, I am very curious on how many people actually finish the courseworks in 1 year, or perhaps, in some form of accelerated manner. I see a number of schools offering incoming students the chance to sit for the re-take exam meant for the previous year students, and if a person had done a master from a decent place or is a transferred student (in which both cases are not exactly that rare), it is highly possible to pass the prelims (especially in micro where almost everywhere's micro sequence tend to be MWG + contract theory at Bolton Dewatripont + game theory at Fudenberg&Tirole standard). If somebody does managed to pass one or two sequences, what does he usually do?? Take the field courses early, or simply enjoy a lighter courseload (and perhaps have a bit more time to start thinking about his research).
Yes -- either is an option. What I've seen people who pass out of one or more core classes do is take one field class first year, along with any of the first year classes they did not test out of. Then, 2nd year, they only have one field class left and can either take additional advanced classes or spend more time on research.

Econ2011
11-22-2010, 04:55 PM
From what I've been told (by a few harvard and MIT graduates), people just don't know where to start, and how to motivate themselves. This is where I was told that having some work experience pays off (if it is in some research related institution). If you have experience going through the whole process, you know what it takes and perhaps, will be a bit easier? Also, having a sound knowledge of office applications (trust me, there is a lot to learn out there which makes your life easier), and statistical analysis software saves you a lot of time.

PS: This is all "I heard it" story. I haven't had a Phd experience yet.

SlowLearner38
11-22-2010, 05:17 PM
I would say that 1st year students being capable of having good research ideas is as rare (or common if you like) as them having the complete set of tools to answer the question at a appropriate level and rigor. I believe it takes time to learn what is a good (or even just relevant) research question or idea; while creativity is important, it is by no means near a sufficient condition to be able to come up with a decent economic question.

First year students usually don't take field courses, so I'd argue that is the reason they don't reveal themselves to have good research questions, not that they aren't capable. It's called a selection problem, but what do I know, I'm only a first year student.

asquare
11-22-2010, 05:24 PM
First year students usually don't take field courses, so I'd argue that is the reason they don't reveal themselves to have good research questions, not that they aren't capable. It's called a selection problem, but what do I know, I'm only a first year student.
I think there's also a treatment effect of the core classes, field classes, and seminars that students attend first and second year. Ideas that you have after that experience will be better than ideas you have before. It's not an issue of capability; it's an issue of experience and training.

mathemagician
11-23-2010, 02:55 AM
But every year, we do see some people being in the job market when they are only in their 4th year (and it really isn't rare) and every now and then, we see people completing even in just 3 years. (And I am not referring to people like Glen Weyl!)

I am just curious to know, where is the part of the program that is possible to be or usually "accelerated"?

People on the job market in 4 is quite rare and most of them are in the top 10 schools. As you move to lower ranked schools, 4 year completion rates are increasingly rare. Many of those who get into the top 10 schools more or less have a year or two of PhD equivalent research and course experience. So on paper it may look like 4 years but they've been at it for a while longer.

econslave
11-24-2010, 05:56 AM
Some types of research take longer than others. Research, in general, takes a lot longer these days. Expectations are higher and the low hanging fruit have already been picked. In addition, advising varies greatly. You can't just go to your advisor every time you have a little question. You have to try to answer as many questions as you can by yourself -- in most cases, approaching your advisor with questions you could look up the answers to yourself will make you look like a fool.

Some people get out in fewer than five, but it's actually pretty rare. It's all a function of how early you're able to start. Some people come in with more preparation than others and need to devote different amounts of time on coursework.

There are a dozen variables, and taking longer isn't a sign of slacking off.

spiderman22
11-24-2010, 06:32 PM
Historically, people tended to finish in 4 years. Most of my professors got their PhDs 4 years after graduating from their BA programs. But, I heard that in the past it was much easier to get placed into good programs. Now people need at least a few publications to get AP positions at the Top 10 schools.

asquare
11-24-2010, 06:53 PM
Historically, people tended to finish in 4 years. Most of my professors got their PhDs 4 years after graduating from their BA programs. But, I heard that in the past it was much easier to get placed into good programs. Now people need at least a few publications to get AP positions at the Top 10 schools.
Not quite. It's true that completion times have gotten longer. However, it is not true that people need even one publication to get a Top 10 job after graduating.

moneyandcredit
11-24-2010, 09:23 PM
Where is this notion coming from that one should have publications before graduation to get a good academic job? Granted, it's not as extreme as thinking one needs to publish as an undergrad to be admitted to a PhD program to begin with (a misconception that I once feared). A PhD program is essentially an academic apprenticeship much like an internship in the public/private sectors. Saying you need to publish while still in school is like saying you need to already be an executive in your junior year to ever land an executive-track job at a Fortune 500 firm.

slightlyconfused1
11-24-2010, 11:49 PM
Where is this notion coming from that one should have publications before graduation to get a good academic job?
In many other fields with quicker publication processes, this is more accurate. For instance, most papers in computer science are published in conference proceedings rather than journals. These papers are shorter and rapidly refereed, and I doubt a computer science Ph.D. without any such publications would do very well on the job market. The publication process in math, though more journal-oriented, is also quicker than in economics, and you see more hotshot students who are already at the research frontier when they enter grad school. Thus I imagine that publications are more important for math Ph.Ds going on the job market than economics students, even though math programs are a year shorter. And in many lab sciences, essentially every grad student contributing in the lab has her name put on the resulting papers, making it almost a given that graduates will have several publications to their name.

Perhaps people with some awareness of the standards in these other fields falsely assume that economics is similar.

Essayvision
11-25-2010, 12:21 AM
Getting your paper published will definitely help especially if it is published on a notable journal. but it is not a prerequisite for landing a job. I know people who got scouted out even before graduation just because they got their paper published on a very notable journal.

Elliephant
11-25-2010, 03:56 AM
Getting your paper published will definitely help especially if it is published on a notable journal. but it is not a prerequisite for landing a job. I know people who got scouted out even before graduation just because they got their paper published on a very notable journal.

Are you speaking out of experience with economists? Because:

In many other fields with quicker publication processes, this is more accurate. [...] Perhaps people with some awareness of the standards in these other fields falsely assume that economics is similar.

Besides, the question wasn't whether good publications help - it is obvious to everyone and his grandmother that they do - but whether they are standard, which, in economics, they're mostly not.

Also, for someone who claims to make a living as a writer/editor, your posts could use better punctuation and correct prepositions.