Should I Get a Ph.D.?

I'm currently pursuing a PhD program here in France. This is a special kind of PhD, called "CIFRE" which roughly means "PhD in a company". You're employed for 3 years by both an academic lab/uni and a company. The goal is to solve an industrial research problem that benefits both the lab and the company.

Personally I'm very happy to be doing this kind of thesis. I'm not in a major lab, so I'm pretty sure I won't be able to fight much against Ivy League PhDs but I'm still getting the degree and I'm okay with it. In the future I don't seek to teach a lot but to do mostly research.

Besides having a good relationship with my advisors I'm also super happy that I don't have a student debt (French education is mostly free -- even for top engineering schools). The pay is good, not as much as an engineer but plenty of people with lower degrees would already be happy with it, so no reason to complain. With this company-linked PhD I also get to study 100% while being officially employed as an engineer, it's on my contract. If I ever want to hide the PhD from my resume (I'm sure I never will), I can basically write "I've been a research engineer for three years" and this would be the truth.

I also get to see how it works in the industrial world and to be more aware when, after the degree, I have to choose between going back to the engineer path or getting further into the academic road.

But the PhD experience is so varied that it's not really the fault of the interviewees/authors! They probably really did have it this good! There is also survivorship bias. And people who have the grit to finish a PhD probably don't want to openly admit weakness on the Internet or, more likely, burn bridges (it's hard to be anonymous about this sort of stuff). Even after they finish the PhD, they still have to worship the ground of their overlords to keep those letters of recommendation flowin'. So we have mostly these rosy happy "learn you a PhD" stories. Not good.

I know you are out there: Young, highly motivated, highly intelligent, unbeatable willpower. You need to know just how bad it can be. You're not getting the whole story.

I'm not talking about "oh no, I might not finish" or "oh gee, maybe it will take me eight years but I'll try real hard and get through." I'm talking real life risks to your mental and physical health, destruction of relationships, opportunity cost, and (potentially) the vaporization of that awesome scientific career that you spent over a decade building because the one person in charge of you with no oversight decided they didn't like you.

I've seen so many amazing, kind, bright, talented, hard-working people exploited for years on end only to be thrown out into the academic garbage can. But the stakes go far beyond academia - what will you do about the panic attacks that continue for years, and years, and years after you finish? How about the insomnia and screaming nightmares? I bet that at least one of your fellow students will wind up in a psychiatric hospital. It could be you.

"Oh come on now," you say. "It's just science! What's so scary about a math problem or writing a few paragraphs. You're either a marshmallow, you're overreacting, or maybe you just didn't have what it takes." The science is the easy part. The hard work is the easy part. It's the people who will rule over you, the people who can (and do) ruin you. These articles always talk about a benevolent best-chum advisor/faculty that you have long conversations with and then go have another cup of tea with. But you never hear about that one narcissist/psychopath on your committee that has done everything in his power to get you out of the program, the micromanager, the manipulator, the grotesque exploitation.

In PhD land, you are at the complete mercy of a very small collection of merciless people who know that you exist to be exploited, and they know that they have you right where they want you. You better hope that those people are benevolent or neutral. In the case of many people I know, this was not the case.

When people write articles about "choosing an advisor" and "how will I know he/she is the one for me," they make it sound like a decision about whether to get a puppy or a kitten. Consider this scenario: You open a dialogue with someone whose work you have studied for years, they offer you a position in their lab, you quit your job and drag your family halfway across the US for this "golden opportunity", and then you find out that this person is by far the biggest jerk that you have ever met and you cannot work with them. What do you do then? Where do you put all of that expertise that you acquired? I've seen this happen over and over again: You start from scratch - time to grind and level up all over again, but now you're not so sure you're good at the thing you're doing - you could be terrible at it.

I could go on and on. Maybe I will someday.

Well, hopefully this will inspire a little more well-rounded picture of what else it can be like to go through with a PhD.

I did my MS and PhD in geology/geophysics and did a good amount of fieldwork, including 8 international field campaigns in places like the Lesser Antilles and Tibet. It's a cool experience to be 23 and send to Nicaragua with a ton of scientific equipment and run small team for a month or two. There are aspects of it that are like tourism, but you go off the backpacker circuit more and interact with the locals, and actually have inescapable intellectual challenges and responsibilities. It's also a bit less heavy than the Peace Corps.

This is pretty common in the earth sciences, although not required. Lots of people in the social sciences have analogous opportunities.

In any case, I think that the overall discussion in the article and in the comments here provide a good range of possible experiences and considerations. But I just wanted to add my piece because it hasn't been mentioned, and it was what really tipped the scales on going to grad school for me. And that was, for me, a great decision.

I'd also add that anyone with an interest in both earth science and coding will find that if they really learn the 'earth' part of the sciences, there are very many opportunities to use relatively simple computations to make advances that a lot of the field scientists haven't worked through yet, and lots of industry opportunities if you're into that as well.

If you are motivated, it is one hell of an experience that you are very unlikely to get anywhere else. This of course depends highly on the group you are joining and the research field that you are going to be in. It is very likely that you get to travel the world and meet interesting new people.

But it is no picknick, and I can confirm some of the other horror stories that you read here. Then again, these made me a better, more focused person.

In my experience the academic world is a closed society of people with similar, often pointless anxieties, an interesting, but quite uniform culture, and often a vague connection with the rest of the universe. I mean, things like publishing, impact factors, tenures and ego-bashing, are things that do not matter in the end, only the science matters, yet many academics' life incessantly revolves around it.

Another place where i have seen a similar "closed world" is the army. People there obsess over mindless pointless things all the time.

I think you should get a PhD because the world is getting richer and you can afford to do this, but make it about learning rather than anything else. If you dont build huge bridges in academia, who cares, you can always join a day job or start your own business. These are exciting times.

[1] http://www.goodreads.com/book/show/460669.Getting_What_You_C...

I also had some of the best times of my life as Ph.D student (and also some of the worst), but it is not something I have regretted doing for one second since.

So, why did I do a PhD? I was told that I would learn lots of new things, and meet lots of new people -- and escape, to some degree, the frustration that I was experiencing with my consulting business.

It's true that I met lots new people. And it's true that I learned a ton. And I'm very proud of the research that I did, resulting in a Web site that is used by thousands of researchers and students every week. (The Modeling Commons -- http://modelingcommons.org/, if you're curious.) And I also go to experience a different sort of frustration than I have had when consulting.

And yet, was it worthwhile? I continue to wrestle with that question. My family and I were hugely stressed for more than a decade. Our finances are improving (thanks to my consulting work), but it'll be another year or so before we're back to where we were. I'm frustrated that I didn't create lots of products and businesses during those 11 years.

And the incredible frustration of the PhD process, and of having an advisor who drove me completely and utterly batty, cannot be ignored.

If you want to do research, then you should do a PhD -- but you should know what you're getting into beforehand, and be really sure that you want to do research.

If you don't want to do research, but want to boost your creds, and if you're single and young, then it might be worthwhile.

If you're like me, in your mid-30s, married, with children, and the primary breadwinner, then you should think long and hard about whether you want the PhD. Several of the others who did PhDs mid-career in my program had spouses earning good incomes, didn't have children, or both.

I do believe that having the PhD has already helped to boost my career a bit, helping me to find newer and bigger clients who somehow think that having a PhD makes you smarter or better than the rest of the population.

But would I recommend it to someone else in my position? Not without a lot of thought and consideration. And an understanding that what you think will take 5 years or so might take much longer than that.

If that person got into an elite Medical, Law, or Business School, they should ask "should I get an MD, JD, or MBA". The reason it's ok to phrase it this way is that attrition rates are typically below one half of one percent in those degree programs at the elite (top 10) level.

Science PhD programs, even elite ones, often have attrition rates at 50%. Engineering is a bit better, at around 35% attrition rates, though this varies by program.

So you should be saying "should I take a 50% shot at getting a PhD" - or, if not, take a very objective look at why the 50% doesn't apply to you. I know, if you got into the PhD program at Berkeley, you justifiably think you're really good at this sort of thing, but really, everyone's good. The 50% who drop or fail out aren't slackers or unintelligent. They're often exceptionally smart and motivated people.

These degrees are very different from professional degrees at the elite level. They are extraordinarily difficult to complete.

On a personal note, I was a PhD student in Industrial Engineering at Berkeley, and I felt there was far too much failure and attrition for such a bright and accomplished group of people. Nobody I knew ended up in an insane asylum (as mentioned in other posts), but it was an emotionally rough experience for many of them, and I could see it getting there. I felt the system did border on being cruel at times.

However, you also learn why nobody has done it before: because those things take a lot of time. And in a PhD program time is not a limited resource, money is. You will be doing stuff that is a waste of time by any objective measure. You have to be very mindful of the time cost of tasks, work, and your choices, because nobody else is. If you're not careful, a very meaningful period of time will have gone by without a lot to show for.

My advice for people who ask me about getting a PhD is that it is risky entering a PhD program without certainty in what you want to do. You can go to college and figure out what to do. But in a doctoral program, you are too likely to get lost in the system, have a bad experience, waste too much time, and accrue too much opportunity cost. You will regret it if that happens.

_______

*PhD in a Physics/Engineering program and research work in neuroscience and medicine. My one reccurring nightmare in life is waking up certain that I'm missing a credit, signature, or form and I'm still in graduate school.

I wish someone had told me that before I started.

You either want to or need to but never should.

Yup, I've seen Ph.D. programs destroy a lot of really good people. People crushed for life, suicide, etc.

I got through okay, but some of the politics was grim. I got into a fight, had to take a year off, and, net, a department Chair and three profs got fired.

How'd I get through? Mostly just did the work on my own. Entered the program very well prepared. Brought my own Ph.D. dissertation research problem, with a good intuitive understanding of how to get a solution, with me to the program and did the real research part independently in my first summer.

Along the way I polished my halo:

One way was in a course, supposed to be really hard. The course was carefully graded, and the intention was that the class be competitive.

But before the course, I'd studied the material, in part in courses but mostly independently, over and over from a stack of the best books, elementary, intermediate, and advanced, applied the material, understood quite a lot about the corresponding numerical analysis, had written software for the material, etc. I could have given all but a few of the lectures on the first day of the class. So, on graded homework, tests, mid-term, and final. I blew away all the other students by wide margins, and I wasn't even trying to be competitive.

In a course there was a question but no answer. So, I asked for reading course as a chance to find an answer. Two weeks later, from working sitting by my wife on our bed as she watched TV, I had a nice, clean answer, with more than I'd hoped to get. Two weeks, course over. Work publishable -- did publish it later.

So, I suggest:

(1) Be very well prepared, from undergraduate school, a Masters from another school, on the job learning, independent study, whatever.

(2) Get an applied Ph.D., say, in engineering or "applied science* or some such. Then for the research, start with a problem from outside academics. Get at least a good intuitive solution before entering the Ph.D. program.

If the lectures, seminars, etc. of the program can give you some tools, ideas, etc. to help you with your research, fine.

(3) Do not ask for a research problem or research direction from a professor. Instead, just do the work independently. If there is any question about the quality of the work, then publish it or at least get it accepted for publication.

The big question, though, is why bother?

One point: Usually a person without a Ph.D. doesn't want to work with a person with a Ph.D. The person without can feel intimidated and threatened, and that one might guess that a Ph.D. might be relevant to his work he can take as an insult. People without have a lot of ways to denigrate people with.

Some of the challenge of a Ph.D. is summarized by D. Knuth in a remark in his The TeXBook:

"The traditional way is to put off all creative aspects until the last part of graduate school. For seventeen or more years, a student is taught examsmanship, then suddenly after passing enough exams in graduate school he's told to do something original."

That "suddenly ... told" can be a big shock.

Here are two common problems:

(1) Good Students.

Ph.D. programs tend to want only good students as in PBK, Summa Cum Laude, Woodrow Wilson, NSF Fellowships, etc.

Well, one of the more common ways to be such a good student is, in addition to being bright, liking the material, being highly determined, and working hard, is to be terribly afraid, of criticism, failure, failing to come up to what parents wanted, what high school teachers expected, of some relative saying that they "expect great things", etc.

So, such a student can be a case of anxiety disease, have done well in K-12 and in college just by making A grades and otherwise not thinking much about anything else. Then the "suddenly ... told" can be a severe challenge, a risk of failure, of the first criticism in life, a threat to self image as a great student, etc.

There can be anxiety, stress, depression, failure, more stress, clinical depression, and suicide.

It can help to have (a) a thick skin and (b) good grounding academically, etc. before entering a Ph.D. program.

I got a thick skin in grades 1-8 -- the teachers treated me like dirt. With my success in math in grades 9-12, I got a little better treatment. But the big day was when the SAT scores came back and the teacher who'd had me in the sixth grade read the Math score and, gulp, "There, uh, there, uh, must be some mistake." Yup, there had been, hers and that of the rest of the teachers, for 12 long, painful, largely wasteful years where I'd been treated like dirt.

But someone up there liked me: I got sent to an NSF summer program in math and physics.

I'd learned that there was no direct way I could ever hope to please the teachers, but with math I could do work that neither they nor anyone else could fault. So, math it was. And physics.

(2) Opaque Criteria.

Mostly students are not told just what the criteria are for the work, the social aspects of the field, the department politics, for research, for publication, for a dissertation, etc.

So, students can do too little in some respects and get into trouble or do too much in other respects, go too slowly, and again get into trouble.

My solution: Largely avoid the opaque stuff. For the learning, do that largely independently. For the research, do that independently and well enough to be publishable, e.g., "new, correct, and significant" and, in case there is any doubt, just submit the work for objective, blind, expert peer-review and, then, publication.

I haven't been at all interested in an academic career or much interested in publishing, but all the papers I've written as sole author or co-author have been accepted for publication right away; so, it's possible to follow the work style I've explained and get published. Getting published in academics is much like making money in business -- lots of other potential problems melt away.

For efforts, work style, such as I have described here, independently and/or in a grad program, is there any benefit? Well, all the best work I've done in my career has been from just that work style. Currently I'm doing a start-up, and the crucial, core secret sauce is from just such work, with some of the grad school work a big source of the prerequisites. So, if my start-up is successful, then the work style I explained above will have been successful.

Disappointing.